Seedance 2.5 API Documentation

Seedance 2.5 is a high-quality AI video generator. This is the standard model — best quality, with output up to 30 seconds at 720p.

What this does: Describe a scene in plain words and get a short video of it.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-5/v1/text-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Text to Video - Seedance 2.5 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-5/v1/text-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [{"type": "text", "text": "A cat walking on the beach"}]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-5/v1/text-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [{"type": "text", "text": "A cat walking on the beach"}]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-5/v1/text-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [{ type: 'text', text: 'A cat walking on the beach' }]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-5/v1/text-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [{"type": "text", "text": "A cat walking on the beach"}]
  }'

Output

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Request Parameters - Seedance 2.5 text-to-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 text item	Input content array. Pass a single text item: `{"type":"text","text":"…"}`.
`content[].type`	Yes	string	—	"text"	Content item type.
`content[].text`	Yes	string	—	Recommended: up to 1,000 English words (or 500 Chinese characters)	Prompt describing the desired video. Seedance 2.5 generates natively from English, Spanish, Indonesian, Portuguese, Japanese, Malay, Thai, Arabic, Vietnamese and Korean prompts. Very long prompts scatter the model’s attention and details start to get dropped.
`tools`	No	array	—	`[{"type":"web_search"}]`	Optional enhancement — text-to-video only.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio. `adaptive` lets the model choose the ratio from your prompt.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution — at `16:9`, `480p` is 854×480 and `720p` is 1280×720. Seedance 2.5 does not support 1080p or 4k.
`duration`	No	integer	`-1`	4–30, or `-1` (model auto-selects)	Output video length in seconds. Use `-1` to let the model choose a suitable whole-second length inside the 4–30s range.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video — speech, sound effects and background music inferred from the prompt and the visuals. Put spoken lines in double quotes inside the prompt for the best result. Generated audio is returned as a 2-channel stereo track, though the stereo field is narrow; set `false` for a silent video.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`output_format`	No	string	"mp4"	"mp4", "mov"	Container for the generated video. `mp4` plays everywhere; `mov` is the professional option (H.264 video, YUV 4:4:4 chroma, PCM audio) for colour grading, keying and compositing, and is recommended when the clip will be fed back in for editing or extension — some players cannot open it.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation (the default 172800 is 48 hours). Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; equal priorities stay FIFO and running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types

Type	Format	Description
text	`{"type":"text","text":"..."}`	The generation prompt (required). May include in-prompt directives below.

In-Prompt Directives (optional)

Seedance 2.5 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "A cinematic shot of a cat on a tropical beach at golden hour --resolution 720p --seed 42 --framepersecond 24 --camerafixed false"}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 15,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-5' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.5 API Pricing — Per-Second Rates

Seedance 2.5 is the newest generation: up to 30-second output, up to 50 multimodal reference assets, native multilingual audio, and video extension. 480p/720p only.

Resolution	Price (USD)
480p	$0.103/sec
720p	$0.231/sec

Seedance 2.5 API Documentation

Seedance 2.5 is a high-quality AI video generator. This is the standard model — best quality, with output up to 30 seconds at 720p.

What this does: Give a starting image (and optionally an ending image), and it creates a video that animates between them.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-5/v1/first-last-frame-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Image to Video (First & Last Frames) - Seedance 2.5 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-5/v1/first-last-frame-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-5/v1/first-last-frame-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-5/v1/first-last-frame-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png' } },
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-5/v1/first-last-frame-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type":"image_url","image_url":{"url":"https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
      {"type":"image_url","image_url":{"url":"https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.5 image-to-video (first & last frames)

Note: For first/last-frame generation, ratio must be adaptive — the output follows the first frame's aspect ratio.

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1–2 `image_url` items; optional 1 `text` item	Input list. One image = first-frame mode; two images = first + last frame mode. `video_url` and `audio_url` items are rejected on this route with `400`.
`content[].type`	Yes	string	—	"text", "image_url"	Content item type.
`content[].text`	No	string	—	Recommended: ≤ 1,000 English words (or 500 Chinese characters)	Optional motion / scene-transition prompt. English works on every model; Seedance 2.5 also accepts Spanish, Portuguese, Japanese, Korean, Indonesian, Malay, Thai, Arabic and Vietnamese. Very long prompts scatter the model's attention and details get dropped — keep it focused.
`content[].image_url.url`	Yes	string	—	Public HTTPS URL, base64 data URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; 300–6000 px per side; aspect ratio (w/h) 0.4–2.5; < 30 MB per image	The frame image. With one image it is the first frame and motion is added from there; with two, the first is the opening frame and the second the closing frame. Base64 must be `data:image/<format>;base64,…` with a lowercase format. Each image must stay under 30 MB and the whole request body under 64 MB — use URLs rather than base64 for large files.
`content[].role`	Required when 2 images	string	`"first_frame"` (single image)	"first_frame", "last_frame"	Frame position. In two-image mode the provider requires `role` on both items; the gateway auto-assigns `first_frame` to the 1st image and `last_frame` to the 2nd in array order if you omit it. The two frames may be identical. If their aspect ratios differ, the first frame takes precedence and the last frame is centre-cropped to match.
`ratio`	No	string	"adaptive"	`"adaptive"` only	Locked for this task type — Seedance 2.5 first/last-frame generation accepts no other value and rejects one with `400`. The output automatically keeps the first frame's aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. Seedance 2.5 does not support 1080p or 4k.
`duration`	No	integer	`-1`	4–30, or `-1` (model auto-selects)	Output video length in seconds. Use `-1` to let the model pick a suitable whole-second length inside the 4–30 s range.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio — voice, sound effects and background music — matched to the prompt and the visuals. Generated audio is returned as a 2-channel stereo track, though the stereo field is narrow. Put spoken lines in double quotes inside the prompt for the best dialogue result.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`output_format`	No	string	"mp4"	"mp4", "mov"	Container for the generated video. `mov` carries higher colour precision and is recommended for video editing and extension; some players cannot open it.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation. Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types

Type	Max	Format / Size	Description
image_url	2	jpeg/png/webp/bmp/tiff/gif/heic/heif · < 30 MB each	1st = first frame (required); 2nd = last frame (optional). `{"type":"image_url","image_url":{"url":"https://..."}}`
text	1	—	Optional motion/style prompt. `{"type":"text","text":"..."}`

In-Prompt Directives (optional)

Seedance 2.5 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Content Item Types

Type	Format	Description
`text`	`{"type":"text","text":"..."}`	Optional. Text prompt describing motion / scene transition. Maximum 1 `text` item.
`image_url` (1st)	`{"type":"image_url","image_url":{"url":"https://..."}}`	Required. First `image_url` in array order. Worker auto-injects role `first_frame`.
`image_url` (2nd)	`{"type":"image_url","image_url":{"url":"https://..."}}`	Required. Second `image_url` in array order. Worker auto-injects role `last_frame`.
`video_url`	—	Not accepted on this route (rejected with 400).
`audio_url`	—	Not accepted on this route (rejected with 400).

Image Constraints (per `image_url` item)

Constraint	Value
Formats	jpeg, png, webp, bmp, tiff, gif, heic, heif
Max file size	Under 30 MB per image; total request body under 64 MB
Pixel dimensions	300–6000 px on each side
Aspect ratio (w/h)	0.4 – 2.5
Number of images	1 (first frame only) or 2 (first + last frame)
How to supply	Public HTTPS URL, `data:image/<format>;base64,…` (lowercase format), or `asset://<ID>`. Prefer URLs for large files.

Example Request

{
  "content": [
    {"type": "text", "text": "Time-lapse from winter to spring; snow melts, leaves green, flowers bloom; cinematic motion."},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "ratio": "adaptive",
  "duration": 5,
  "resolution": "720p",
  "generate_audio": true,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-5' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.5 API Pricing — Per-Second Rates

Seedance 2.5 is the newest generation: up to 30-second output, up to 50 multimodal reference assets, native multilingual audio, and video extension. 480p/720p only.

Resolution	Price (USD)
480p	$0.103/sec
720p	$0.231/sec

Seedance 2.5 API Documentation

Seedance 2.5 is a high-quality AI video generator. This is the standard model — best quality, with output up to 30 seconds at 720p.

What this does: Give it photos (and/or short clips) of people or things, and it creates a video featuring them.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-5/v1/reference-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Reference to Video - Seedance 2.5 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-5/v1/reference-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-5/v1/reference-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-5/v1/reference-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-5/v1/reference-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.5 reference-to-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	optional text + 0–30 images and/or 0–10 videos and/or 0–10 audio clips; 50 references max	Multimodal reference input. Any combination of image, video and audio references — in 2.5 audio alone is a valid reference. Keep the whole request body under 64 MB; send large assets as URLs rather than base64.
`content[].type`	Yes	string	—	"text", "image_url", "video_url", "audio_url"	Content item type.
`content[].text`	No	string	—	Recommended: ≤ 1,000 English words (≤ 500 Chinese characters)	Prompt; refer to uploads by order, e.g. "Image 1", "Video 1". English is supported by every Seedance model; 2.5 also generates natively from Spanish, Portuguese, Japanese, Korean, Indonesian, Malay, Thai, Arabic and Vietnamese prompts. Longer prompts scatter the model's attention and details start getting dropped.
`content[].image_url.url`	Conditional	string	—	URL, base64 data URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; max 30 MB each; 300–6000 px per side; aspect ratio (w/h) 0.4–2.5; 1–30 images	Upload up to 30 reference images the model takes subject, style or composition cues from, to keep your result consistent. JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC or HEIF; under 30 MB each; every side 300–6000 px with a width/height ratio between 0.4 and 2.5.
`content[].video_url.url`	Conditional	string	—	URL or `asset://<ID>`; mp4/mov; max 200 MB each; 2–30 s per clip and 30 s combined; 300–6000 px per side; aspect ratio (w/h) 0.4–2.5; 409,600–8,295,044 total pixels; 24–60 FPS	Upload up to 10 reference videos the model takes motion or style cues from. MP4 or MOV, 480p to 4K source; each clip 2–30 seconds and under 200 MB; 30 seconds combined across all videos. Clips under 20 seconds usually transfer motion best.
`content[].audio_url.url`	Conditional	string	—	URL, base64 data URI, or `asset://<ID>`; wav/mp3; max 15 MB each; 2–30 s per clip and 30 s combined	Upload up to 10 audio clips as a reference for music, dialogue, voice, tone or timbre. MP3 or WAV; each clip 2–30 seconds and under 15 MB; 30 seconds combined. In Seedance 2.5 audio can be the only reference you send.
`content[].role`	Required on refs	string	—	"reference_image", "reference_video", "reference_audio"	Role of each reference item — `reference_image` on images, `reference_video` on videos, `reference_audio` on audio. Required on every non-text item. (The provider's `first_frame`/`last_frame` roles belong to the separate image-to-video modes, which cannot be mixed with reference input.)
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio. `adaptive` lets the model pick one from the task type and your inputs. Video-editing and video-extension requests keep the input video's ratio and accept only `adaptive`.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. Seedance 2.5 does not support 1080p or 4k.
`duration`	No	integer	`-1`	4–30, or `-1` (model auto-selects)	Output video length in seconds. Use `-1` to let the model choose a suitable length. Video-editing requests accept only `-1` and keep the input clip's length.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video — speech, sound effects and music matched to the prompt and footage. Put spoken lines in double quotes to steer the dialogue. Generated audio is returned as a 2-channel stereo track, though the stereo field is narrow.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`output_format`	No	string	"mp4"	"mp4", "mov"	Container for the generated video. `mov` carries higher colour precision and is recommended for video editing and extension; some players cannot open it.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation. Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types & Limits

Type	Max count	Format / Size	Description
image_url	up to 30	JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC, HEIF · < 30 MB each · 300–6000 px/side	Reference image for character consistency, style, or composition. `{"type":"image_url","image_url":{"url":"https://..."}}`
video_url	up to 10	MP4, MOV · ≤ 200 MB each · 2–30s per clip · ≤ 30s combined	Reference video for motion / style transfer. `{"type":"video_url","video_url":{"url":"https://..."}}`
audio_url	up to 10	MP3, WAV · 2–30s per clip · ≤ 30s combined · ≤ 15 MB each	Reference audio for audio-driven generation & lip-sync. In Seedance 2.5 audio can be the only reference. `{"type":"audio_url","audio_url":{"url":"https://..."}}`
text	1	—	Prompt. Address references by their 1-indexed array order — `[Image1]`, `[Video1]`, `[Audio1]` (plain "Image 1" also works).

Asset budget: up to 50 references in total — at most 30 images, 10 videos and 10 audio clips. Video references may total 30 seconds combined, and audio references 30 seconds combined.

In-Prompt Directives (optional)

Seedance 2.5 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Use Image 1 as the opening frame and Image 2 as the closing frame. Smooth cinematic transition between them with gentle camera motion. --resolution 720p"},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 11,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-5' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.5 API Pricing — Per-Second Rates

Seedance 2.5 is the newest generation: up to 30-second output, up to 50 multimodal reference assets, native multilingual audio, and video extension. 480p/720p only.

Resolution	Price (USD)
480p	$0.103/sec
720p	$0.231/sec

Seedance 2.5 API Documentation

Seedance 2.5 is a high-quality AI video generator. This is the standard model — best quality, with output up to 30 seconds at 720p.

What this does: Upload a video plus a text instruction, and it changes the video to match your request.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-5/v1/edit-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Edit Video - Seedance 2.5 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-5/v1/edit-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "text", "text": "Make truck Orange"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-5/v1/edit-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "text", "text": "Make truck Orange"},
        {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-5/v1/edit-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'text', text: 'Make truck Orange' },
    { type: 'video_url', video_url: { url: 'https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-5/v1/edit-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "text", "text": "Make truck Orange"},
      {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.5 edit-video

Note: For video editing, ratio must be adaptive and duration must be -1 — the output tracks the input video.

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 text + 1 video_url (required) + optional image_url	Edit instruction + the video to edit, plus an optional reference image.
`content[].type`	Yes	string	—	"text", "image_url", "video_url"	Content item type.
`content[].text`	Yes	string	—	Recommended: under 1000 words	Edit instruction; refer to uploads by order, e.g. "Image 1", "Video 1". Include an explicit edit verb (add, remove, replace, change, edit) so the model classifies the request as an edit instead of a new generation.
`content[].image_url.url`	No	string	—	URL, base64 URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; max 30 MB; 300–6000 px per side; aspect ratio 0.4–2.5	Optional reference image — the object, character, or scene to bring into the edited video. Under 30 MB. Omit it to edit the video from the text instruction alone.
`content[].video_url.url`	Yes	string	—	URL or `asset://<ID>`; mp4/mov (H.264 or H.265 video, AAC or MP3 audio); 4–30 s; 24–60 FPS; max 200 MB; 300–6000 px per side	Upload the video you want to edit. MP4 or MOV, up to 200 MB, and 4–30 seconds long — a source clip shorter than 4 s or longer than 30 s is rejected. MOV is recommended for editing (higher colour precision).
`content[].role`	No	string	auto-assigned	"reference_image", "reference_video"	Asset role. The gateway sets it automatically from the item type; any value you send is overwritten.
`ratio`	No	string	"adaptive"	"adaptive" only	Locked for video editing. The output always keeps the source video’s aspect ratio, so `adaptive` is the only accepted value.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. Seedance 2.5 does not support 1080p or 4k.
`duration`	No	integer	`-1`	`-1` only	Locked for video editing. The output tracks the source video’s length (it may come back up to 0.4 s shorter, and may be a non-integer number of seconds), so `-1` is the only accepted value.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`output_format`	No	string	"mp4"	"mp4", "mov"	Container for the generated video. `mov` carries higher colour precision and is recommended for video editing and extension; some players cannot open it.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation. Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types

Type	Max	Format / Size	Description
video_url	10	MP4 or MOV · ≤ 200 MB each · 4–30 s · 30 s total	Source video to edit (required). `{"type":"video_url","video_url":{"url":"https://..."}}`
text	1	—	Edit instructions. `{"type":"text","text":"..."}`
image_url	30	JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC, HEIF · < 30 MB each	Optional style / reference image. `{"type":"image_url","image_url":{"url":"https://..."}}`

In-Prompt Directives (optional)

Seedance 2.5 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--rs`	`--rs 720p`	Output resolution (`480p` / `720p`). Equivalent to the `resolution` field.
`--wm`	`--wm true`	Add the "AI Generated" watermark. Equivalent to the `watermark` field.

Example Request

{
  "content": [
    {"type": "text", "text": "Restyle the source video using Image 1 color palette, framing Image 2 as the final beat. --resolution 720p"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "adaptive",
  "duration": -1,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-5' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.5 API Pricing — Per-Second Rates

Seedance 2.5 is the newest generation: up to 30-second output, up to 50 multimodal reference assets, native multilingual audio, and video extension. 480p/720p only.

Resolution	Price (USD)
480p	$0.103/sec
720p	$0.231/sec

Seedance 2.5 API Documentation

Seedance 2.5 is a high-quality AI video generator. This is the standard model — best quality, with output up to 30 seconds at 720p.

What this does: Upload a video plus a text instruction describing what should happen next, and it generates a continuation of that clip.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-5/v1/extend-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Extend Video - Seedance 2.5 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-5/v1/extend-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "text", "text": "Extend Video 1 forward: the truck keeps driving down the coast road as the sun sets behind it."},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-5/v1/extend-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "text", "text": "Extend Video 1 forward: the truck keeps driving down the coast road as the sun sets behind it."},
        {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-5/v1/extend-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'text', text: 'Extend Video 1 forward: the truck keeps driving down the coast road as the sun sets behind it.' },
    { type: 'video_url', video_url: { url: 'https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-5/v1/extend-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "text", "text": "Extend Video 1 forward: the truck keeps driving down the coast road as the sun sets behind it."},
      {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.5 extend-video

Note: For video extension, ratio must be adaptive; duration may be 4–30 or -1.

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 text + 1 video_url (required), plus optional extra references — up to 50 assets total (30 images, 10 videos, 10 audio)	The extension instruction plus the clip you want to continue. Extra images, videos or audio can be attached as references.
`content[].type`	Yes	string	—	"text", "image_url", "video_url", "audio_url"	Content item type.
`content[].text`	Yes	string	—	Recommended: under 1000 words	Describes what happens in the continuation. Must contain an extension cue such as "extend forward", "extend backward", "continue" or "continue the story" — without one the model may classify the job as an edit instead and reject it. Refer to uploads by order, e.g. "Video 1", "Image 1".
`content[].video_url.url`	Yes	string	—	URL or `asset://<ID>`; mp4/mov; max 200 MB; 2–30 s; 24–60 FPS; 300–6000 px per side; aspect ratio 0.4–2.5	The clip you want to extend. Accepts `mp4` or `mov` — `mov` is recommended for the best colour and audio continuity across the join. Up to 200 MB and 2–30 seconds long. Extension can run forwards or backwards from this clip.
`content[].image_url.url`	No	string	—	URL, base64 URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; max 30 MB; 300–6000 px per side; up to 30 images	Optional reference image — an object, character or scene to carry into the continuation.
`content[].audio_url.url`	No	string	—	URL, base64 URI, or `asset://<ID>`; wav/mp3; max 15 MB; 2–30 s; up to 10 clips (30 s combined)	Optional reference audio — voice, music or sound effects to guide the continuation’s soundtrack.
`content[].role`	Conditional	string	—	"reference_video", "reference_image", "reference_audio", "first_frame", "last_frame"	Role of the attached asset. Videos use `reference_video` and audio uses `reference_audio`; images use `reference_image`, or `first_frame`/`last_frame` to pin an image to the start or end.
`ratio`	No	string	"adaptive"	"adaptive" only	Output aspect ratio. Locked for video extension — the result always keeps the aspect ratio of the clip being extended, so `adaptive` is the only accepted value. Passing 16:9, 9:16 or any other ratio fails after the job is queued.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. Seedance 2.5 does not support 1080p or 4k.
`duration`	No	integer	`-1`	4–30, or `-1` (model auto-selects)	Length of the generated video in seconds. Use `-1` to let the model pick a suitable whole-second length in the 4–30 range. The source clip has its own separate limit (2–30 s) and counts toward the 30 s reference-video budget.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`output_format`	No	string	"mp4"	"mp4", "mov"	Container for the generated video. `mov` carries higher colour precision and is recommended for extension — ideally for both the source clip and the output, which gives the most seamless join. Some players cannot open `mov`.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation. Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types

Type	Max	Format / Size	Description
video_url	1 source (10 videos total)	MP4 or MOV · ≤ 200 MB · 2–30 s · 24–60 FPS	Source video to extend (required). `{"type":"video_url","video_url":{"url":"https://..."}}`
text	1	—	Extension instructions — must include an extension cue such as "continue" or "extend forward". `{"type":"text","text":"..."}`
image_url	30	JPG, PNG, WEBP · ≤ 30 MB	Optional style / reference image. `{"type":"image_url","image_url":{"url":"https://..."}}`
audio_url	10	WAV or MP3 · ≤ 15 MB · 2–30 s	Optional reference audio. `{"type":"audio_url","audio_url":{"url":"https://..."}}`

In-Prompt Directives (optional)

Seedance 2.5 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Continue Video 1: the camera pulls back to reveal the whole valley, keeping the colour palette of Image 1 and ending on the composition in Image 2. --resolution 720p"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "adaptive",
  "duration": 11,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-5' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-5",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-5_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-5",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-5_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.5 API Pricing — Per-Second Rates

Seedance 2.5 is the newest generation: up to 30-second output, up to 50 multimodal reference assets, native multilingual audio, and video extension. 480p/720p only.

Resolution	Price (USD)
480p	$0.103/sec
720p	$0.231/sec

Seedance 2.0 Mini API Documentation

Seedance 2.0 Mini is the most affordable version of Seedance 2.0 — the cheapest way to create a video (480p and 720p).

What this does: Describe a scene in plain words and get a short video of it.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0-mini/text-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Text to Video - Seedance 2.0 Mini API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0-mini/text-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [{"type": "text", "text": "A cat walking on the beach"}]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0-mini/text-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [{"type": "text", "text": "A cat walking on the beach"}]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0-mini/text-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [{ type: 'text', text: 'A cat walking on the beach' }]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0-mini/text-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [{"type": "text", "text": "A cat walking on the beach"}]
  }'

Output

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-mini",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-mini",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Request Parameters - Seedance 2.0 Mini text-to-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1+ text item	Input content array. Pass a single text item: `{"type":"text","text":"…"}`.
`content[].type`	Yes	string	—	"text"	Content item type.
`content[].text`	Yes	string	—	Recommended: under 1000 words	Prompt describing the desired video. English, Japanese, Indonesian, Spanish, and Portuguese prompts are supported.
`tools`	No	array	—	`[{"type":"web_search"}]`	Optional enhancement — text-to-video only.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. The Mini variant supports 480p and 720p only — 1080p and 4k are not available.
`duration`	No	integer	5	4–15, or `-1` (auto)	Output video length in seconds. Set -1 to let the model pick the best length automatically.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add a watermark to the output video.

Billing is per second of actual output video, rounded up to the next whole second — clips typically run ~0.1s over the requested duration, so a 5-second request usually bills 6 seconds. The JSON resolution field takes precedence over a --resolution text directive in the prompt; if omitted, the default is 720p (billed at the 720p rate). Set "resolution": "480p" in the JSON body for the 480p rate.

Content Item Types

Type	Format	Description
text	`{"type":"text","text":"..."}`	The generation prompt (required). May include in-prompt directives below.

In-Prompt Directives (optional)

Seedance 2.0 Mini also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "A cinematic shot of a cat on a tropical beach at golden hour --resolution 720p --seed 42 --framepersecond 24 --camerafixed false"}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 15,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0-mini' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-mini",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-mini",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-mini_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 Mini API Pricing — Per-Second Rates

Seedance 2.0 is the production model. Mini is the cheapest tier; Fast trades quality for speed. 1.0 Pro / 1.0 Lite are legacy variants.

Resolution	Price (USD)
480p	~~$0.035/sec~~$0.014/sec
720p	~~$0.0756/sec~~$0.0302/sec

Seedance 2.0 Mini API Documentation

Seedance 2.0 Mini is the most affordable version of Seedance 2.0 — the cheapest way to create a video (480p and 720p).

What this does: Give a starting image (and optionally an ending image), and it creates a video that animates between them.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0-mini/first-last-frame-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Image to Video (First & Last Frames) - Seedance 2.0 Mini API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0-mini/first-last-frame-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0-mini/first-last-frame-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0-mini/first-last-frame-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png' } },
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0-mini/first-last-frame-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type":"image_url","image_url":{"url":"https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
      {"type":"image_url","image_url":{"url":"https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-mini",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-mini",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 Mini image-to-video (first & last frames)

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 or 2 image_url; optional text	First-frame (1 image) or first+last-frame (2 images) mode.
`content[].type`	Yes	string	—	"text", "image_url"	Content item type.
`content[].text`	No	string	—	Recommended: under 1000 words	Optional motion guidance prompt.
`content[].image_url.url`	Yes	string	—	URL, base64 URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; under 30 MB; 300–6000 px per side	Upload the image your video starts from — it becomes the first frame, then motion is added. JPG, PNG or WebP; under 30 MB.
`content[].role`	Required when 2 images	string	`"first_frame"` (single image)	"first_frame", "last_frame"	Frame position. Optional for a single image (treated as first_frame); in two-frame mode both images need a role — first_frame on the first, last_frame on the second.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. The Mini variant supports 480p and 720p only — 1080p and 4k are not available.
`duration`	No	integer	5	4–15, or `-1` (auto)	Output video length in seconds. Set -1 to let the model pick the best length automatically.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add a watermark to the output video.

Billing is per second of actual output video, rounded up to the next whole second — clips typically run ~0.1s over the requested duration, so a 5-second request usually bills 6 seconds. The JSON resolution field takes precedence over a --resolution text directive in the prompt; if omitted, the default is 720p (billed at the 720p rate). Set "resolution": "480p" in the JSON body for the 480p rate.

Content Item Types

Type	Max	Format / Size	Description
image_url	2	JPG, PNG, WEBP · < 30 MB each	1st = first frame (required); 2nd = last frame (optional). `{"type":"image_url","image_url":{"url":"https://..."}}`
text	1	—	Optional motion/style prompt. `{"type":"text","text":"..."}`

In-Prompt Directives (optional)

Seedance 2.0 Mini also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Content Item Types

Type	Format	Description
`text`	`{"type":"text","text":"..."}`	Optional. Text prompt describing motion / scene transition. Maximum 1 `text` item.
`image_url` (1st)	`{"type":"image_url","image_url":{"url":"https://..."}}`	Required. First `image_url` in array order. Worker auto-injects role `first_frame`.
`image_url` (2nd)	`{"type":"image_url","image_url":{"url":"https://..."}}`	Required. Second `image_url` in array order. Worker auto-injects role `last_frame`.
`video_url`	—	Not accepted on this route (rejected with 400).
`audio_url`	—	Not accepted on this route (rejected with 400).

Image Constraints (per `image_url` item)

Constraint	Value
Format	The URL of the starting frame image to animate. Supported formats: JPEG, PNG, WebP. Max 30 MB.

Example Request

{
  "content": [
    {"type": "text", "text": "Time-lapse from winter to spring; snow melts, leaves green, flowers bloom; cinematic motion."},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "ratio": "adaptive",
  "duration": 5,
  "resolution": "720p",
  "generate_audio": true,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0-mini' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-mini",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-mini",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-mini_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 Mini API Pricing — Per-Second Rates

Resolution	Price (USD)
480p	~~$0.035/sec~~$0.014/sec
720p	~~$0.0756/sec~~$0.0302/sec

Seedance 2.0 Mini API Documentation

Seedance 2.0 Mini is the most affordable version of Seedance 2.0 — the cheapest way to create a video (480p and 720p).

What this does: Give it photos (and/or short clips) of people or things, and it creates a video featuring them.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0-mini/reference-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Reference to Video - Seedance 2.0 Mini API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0-mini/reference-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0-mini/reference-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0-mini/reference-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0-mini/reference-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-mini",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-mini",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 Mini reference-to-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	optional text + at least one image_url or video_url (audio_url cannot be the only reference); ordered text → images → videos → audio	Multimodal reference input.
`content[].type`	Yes	string	—	"text", "image_url", "video_url", "audio_url"	Content item type.
`content[].text`	No	string	—	Recommended: under 1000 words; may include `@image1` / `@video1` / `@audio1` refs	Prompt with optional reference handles.
`content[].image_url.url`	Conditional	string	—	URL, base64 data URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; under 30 MB each	Upload up to 9 reference images the model takes style or subject cues from, to keep your result consistent. JPG, PNG or WebP; under 30 MB each.
`content[].video_url.url`	Conditional	string	—	URL or `asset://<ID>`; mp4/mov; max 200 MB each; 2–15 s each, 15 s combined	Upload up to 3 reference videos the model takes motion or style cues from. MP4; under 50 MB each; up to 15 seconds total.
`content[].audio_url.url`	Conditional	string	—	URL, base64 URI, or `asset://<ID>`; wav/mp3; max 15 MB each; 2–15 s each, 15 s combined	Upload up to 3 audio clips as a reference. MP3 or WAV; under 15 MB each; up to 15 seconds total.
`content[].role`	Required on refs	string	—	"reference_image", "reference_video", "reference_audio"	Asset role designation.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. The Mini variant supports 480p and 720p only — 1080p and 4k are not available.
`duration`	No	integer	5	4–15, or `-1` (auto)	Output video length in seconds. Set -1 to let the model pick the best length automatically.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add a watermark to the output video.

Billing is per second of actual output video, rounded up to the next whole second — clips typically run ~0.1s over the requested duration, so a 5-second request usually bills 6 seconds. The JSON resolution field takes precedence over a --resolution text directive in the prompt; if omitted, the default is 720p (billed at the 720p rate). Set "resolution": "480p" in the JSON body for the 480p rate.

Content Item Types & Limits

Type	Max count	Format / Size	Description
image_url	up to 9	JPG, PNG, WEBP · < 30 MB each	Reference image for character consistency, style, or composition. `{"type":"image_url","image_url":{"url":"https://..."}}`
video_url	up to 3	MP4 · < 50 MB each · ≤ 15s combined	Reference video for motion / style transfer. `{"type":"video_url","video_url":{"url":"https://..."}}`
audio_url	up to 3	MP3, WAV · ≤ 15s combined · < 15 MB	Reference audio for audio-driven generation & lip-sync. Requires at least one image or video reference. `{"type":"audio_url","audio_url":{"url":"https://..."}}`
text	1	—	Prompt. Address references by their 1-indexed array order — `[Image1]`, `[Video1]`, `[Audio1]` (plain "Image 1" also works).

The rule of 12: total references across all types may not exceed 12 (e.g. 9 images + 3 videos = 12, leaving no room for audio). Per-type caps (9 / 3 / 3) always apply within that budget.

In-Prompt Directives (optional)

Seedance 2.0 Mini also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Use Image 1 as the opening frame and Image 2 as the closing frame. Smooth cinematic transition between them with gentle camera motion. --resolution 720p"},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 11,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0-mini' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-mini",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-mini",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-mini_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 Mini API Pricing — Per-Second Rates

Resolution	Price (USD)
480p	~~$0.0478/sec~~$0.0191/sec
720p	~~$0.1074/sec~~$0.043/sec

Seedance 2.0 Mini API Documentation

Seedance 2.0 Mini is the most affordable version of Seedance 2.0 — the cheapest way to create a video (480p and 720p).

What this does: Upload a video plus a text instruction, and it changes the video to match your request.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0-mini/edit-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Edit Video - Seedance 2.0 Mini API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0-mini/edit-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "text", "text": "Make truck Orange"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0-mini/edit-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "text", "text": "Make truck Orange"},
        {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0-mini/edit-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'text', text: 'Make truck Orange' },
    { type: 'video_url', video_url: { url: 'https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0-mini/edit-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "text", "text": "Make truck Orange"},
      {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-mini",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-mini",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 Mini edit-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 text + 1 image_url + 1 video_url	Edit instruction + reference object image + video to edit.
`content[].type`	Yes	string	—	"text", "image_url", "video_url"	Content item type.
`content[].text`	Yes	string	—	Recommended: under 1000 words; may use `@image1` / `@video1`	Edit instruction with reference handles.
`content[].image_url.url`	Yes	string	—	URL, base64 URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; under 30 MB	Upload an image of the object, character, or scene to bring into the edited video. JPG, PNG or WebP; under 30 MB.
`content[].video_url.url`	Yes	string	—	URL or `asset://<ID>`; mp4/mov; max 200 MB; 2–15 s	Upload the video you want to edit. MP4; under 50 MB; 2–15 seconds.
`content[].role`	No	string	auto-assigned	"reference_image", "reference_video"	Asset role. The gateway sets it automatically from the item type; any value you send is overwritten.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. The Mini variant supports 480p and 720p only — 1080p and 4k are not available.
`duration`	No	integer	5	4–15, or `-1` (auto)	Output video length in seconds. Set -1 to let the model pick the best length automatically.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add a watermark to the output video.

Billing is per second of actual output video, rounded up to the next whole second — clips typically run ~0.1s over the requested duration, so a 5-second request usually bills 6 seconds. The JSON resolution field takes precedence over a --resolution text directive in the prompt; if omitted, the default is 720p (billed at the 720p rate). Set "resolution": "480p" in the JSON body for the 480p rate.

Content Item Types

Type	Max	Format / Size	Description
video_url	1	MP4 · < 50 MB	Source video to edit (required). `{"type":"video_url","video_url":{"url":"https://..."}}`
text	1	—	Edit instructions. `{"type":"text","text":"..."}`
image_url	1	JPG, PNG, WEBP · < 30 MB	Optional style / reference image. `{"type":"image_url","image_url":{"url":"https://..."}}`

In-Prompt Directives (optional)

Seedance 2.0 Mini also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Restyle the source video using Image 1 color palette, framing Image 2 as the final beat. --resolution 720p"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 11,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0-mini' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-mini",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0-mini_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-mini",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-mini_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 Mini API Pricing — Per-Second Rates

Resolution	Price (USD)
480p	~~$0.0478/sec~~$0.0191/sec
720p	~~$0.1074/sec~~$0.043/sec

Seedance 2.0 API Documentation

Seedance 2.0 is a high-quality AI video generator. This is the standard model — best quality, with output up to 4K.

What this does: Describe a scene in plain words and get a short video of it.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0/v1/text-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Text to Video - Seedance 2.0 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0/v1/text-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [{"type": "text", "text": "A cat walking on the beach"}]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0/v1/text-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [{"type": "text", "text": "A cat walking on the beach"}]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0/v1/text-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [{ type: 'text', text: 'A cat walking on the beach' }]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0/v1/text-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [{"type": "text", "text": "A cat walking on the beach"}]
  }'

Output

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Request Parameters - Seedance 2.0 text-to-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 text item	Input content array. Pass a single text item: `{"type":"text","text":"…"}`.
`content[].type`	Yes	string	—	"text"	Content item type.
`content[].text`	Yes	string	—	Recommended: under 1000 words	Prompt describing the desired video. English, Japanese, Indonesian, Spanish and Portuguese are supported.
`tools`	No	array	—	`[{"type":"web_search"}]`	Optional enhancement — text-to-video only.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p", "1080p", "4k"	Output resolution. Higher resolutions take longer to render. 4k output is 10-bit H.265 (HEVC).
`duration`	No	integer	5	4–15, or `-1` (model auto-selects)	Output video length in seconds. Use -1 to let the model pick a suitable length.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation. Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types

Type	Format	Description
text	`{"type":"text","text":"..."}`	The generation prompt (required). May include in-prompt directives below.

In-Prompt Directives (optional)

Seedance 2.0 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 1080p`	Output resolution (`480p` / `720p` / `1080p` / `4k`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "A cinematic shot of a cat on a tropical beach at golden hour --resolution 1080p --seed 42 --framepersecond 24 --camerafixed false"}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 15,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 API Pricing — Per-Second Rates

Seedance 2.0 is the production model. Mini is the cheapest tier; Fast trades quality for speed. 1.0 Pro / 1.0 Lite are legacy variants.

Resolution	Price (USD)
480p	$0.07/sec
720p	$0.1512/sec
1080p	$0.3742/sec
4k	$0.7776/sec

Seedance 2.0 API Documentation

Seedance 2.0 is a high-quality AI video generator. This is the standard model — best quality, with output up to 4K.

What this does: Give a starting image (and optionally an ending image), and it creates a video that animates between them.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0/v1/first-last-frame-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Image to Video (First & Last Frames) - Seedance 2.0 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0/v1/first-last-frame-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0/v1/first-last-frame-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0/v1/first-last-frame-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png' } },
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0/v1/first-last-frame-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type":"image_url","image_url":{"url":"https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
      {"type":"image_url","image_url":{"url":"https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 image-to-video (first & last frames)

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 or 2 image_url; optional text	First-frame (1 image) or first+last-frame (2 images) mode.
`content[].type`	Yes	string	—	"text", "image_url"	Content item type.
`content[].text`	No	string	—	Recommended: under 1000 words	Optional motion guidance prompt.
`content[].image_url.url`	Yes	string	—	URL, base64 URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; max 30 MB; 300–6000 px per side	Upload the image your video starts from — it becomes the first frame, then motion is added. JPG, PNG or WebP; under 30 MB.
`content[].role`	Required when 2 images	string	`"first_frame"` (single image)	"first_frame", "last_frame"	Frame position. Required on the 2nd image in two-frame mode. If the two frames have different aspect ratios, the first frame takes precedence and the last frame is center-cropped.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p", "1080p", "4k"	Output resolution. Higher resolutions take longer to render. 4k output is 10-bit H.265 (HEVC).
`duration`	No	integer	5	4–15, or `-1` (model auto-selects)	Output video length in seconds. Use -1 to let the model pick a suitable length.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation. Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types

Type	Max	Format / Size	Description
image_url	2	JPG, PNG, WEBP · < 30 MB each	1st = first frame (required); 2nd = last frame (optional). `{"type":"image_url","image_url":{"url":"https://..."}}`
text	1	—	Optional motion/style prompt. `{"type":"text","text":"..."}`

In-Prompt Directives (optional)

Seedance 2.0 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 1080p`	Output resolution (`480p` / `720p` / `1080p` / `4k`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Content Item Types

Type	Format	Description
`text`	`{"type":"text","text":"..."}`	Optional. Text prompt describing motion / scene transition. Maximum 1 `text` item.
`image_url` (1st)	`{"type":"image_url","image_url":{"url":"https://..."}}`	Required. First `image_url` in array order. Worker auto-injects role `first_frame`.
`image_url` (2nd)	`{"type":"image_url","image_url":{"url":"https://..."}}`	Required. Second `image_url` in array order. Worker auto-injects role `last_frame`.
`video_url`	—	Not accepted on this route (rejected with 400).
`audio_url`	—	Not accepted on this route (rejected with 400).

Image Constraints (per `image_url` item)

Constraint	Value
Format	The URL of the starting frame image to animate. Supported formats: JPEG, PNG, WebP. Max 30 MB.

Example Request

{
  "content": [
    {"type": "text", "text": "Time-lapse from winter to spring; snow melts, leaves green, flowers bloom; cinematic motion."},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "ratio": "adaptive",
  "duration": 5,
  "resolution": "1080p",
  "generate_audio": true,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 API Pricing — Per-Second Rates

Resolution	Price (USD)
480p	$0.07/sec
720p	$0.1512/sec
1080p	$0.3742/sec
4k	$0.7776/sec

Seedance 2.0 API Documentation

Seedance 2.0 is a high-quality AI video generator. This is the standard model — best quality, with output up to 4K.

What this does: Give it photos (and/or short clips) of people or things, and it creates a video featuring them.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0/v1/reference-to-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Reference to Video - Seedance 2.0 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0/v1/reference-to-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0/v1/reference-to-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0/v1/reference-to-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0/v1/reference-to-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 reference-to-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	optional text + 1–9 images and/or 1–3 videos + 0–3 audio (audio cannot be the only reference)	Multimodal reference input.
`content[].type`	Yes	string	—	"text", "image_url", "video_url", "audio_url"	Content item type.
`content[].text`	No	string	—	Recommended: under 1000 words	Prompt; refer to uploads by order, e.g. "Image 1", "Video 1".
`content[].image_url.url`	Conditional	string	—	URL, base64 data URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; max 30 MB; 300–6000 px per side	Upload up to 9 reference images the model takes style or subject cues from, to keep your result consistent. JPG, PNG or WebP; under 30 MB each.
`content[].video_url.url`	Conditional	string	—	URL or `asset://<ID>`; mp4/mov; max 200 MB each; 2–15 s each, combined max 15 s; 24–60 FPS	Upload up to 3 reference videos the model takes motion or style cues from. MP4; under 50 MB each; up to 15 seconds total.
`content[].audio_url.url`	Conditional	string	—	URL, base64 URI, or `asset://<ID>`; wav/mp3; max 15 MB each; 2–15 s each, combined max 15 s	Upload up to 3 audio clips as a reference. MP3 or WAV; under 15 MB each; up to 15 seconds total.
`content[].role`	Required on refs	string	—	"reference_image", "reference_video", "reference_audio"	Asset role designation.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p", "1080p", "4k"	Output resolution. Higher resolutions take longer to render. 4k output is 10-bit H.265 (HEVC).
`duration`	No	integer	5	4–15, or `-1` (model auto-selects)	Output video length in seconds. Use -1 to let the model pick a suitable length.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation. Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types & Limits

Type	Max count	Format / Size	Description
image_url	up to 9	JPG, PNG, WEBP · < 30 MB each	Reference image for character consistency, style, or composition. `{"type":"image_url","image_url":{"url":"https://..."}}`
video_url	up to 3	MP4 · < 50 MB each · ≤ 15s combined	Reference video for motion / style transfer. `{"type":"video_url","video_url":{"url":"https://..."}}`
audio_url	up to 3	MP3, WAV · ≤ 15s combined · < 15 MB	Reference audio for audio-driven generation & lip-sync. Requires at least one image or video reference. `{"type":"audio_url","audio_url":{"url":"https://..."}}`
text	1	—	Prompt. Address references by their 1-indexed array order — `[Image1]`, `[Video1]`, `[Audio1]` (plain "Image 1" also works).

The rule of 12: total references across all types may not exceed 12 (e.g. 9 images + 3 videos = 12, leaving no room for audio). Per-type caps (9 / 3 / 3) always apply within that budget.

In-Prompt Directives (optional)

Seedance 2.0 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 1080p`	Output resolution (`480p` / `720p` / `1080p` / `4k`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Use Image 1 as the opening frame and Image 2 as the closing frame. Smooth cinematic transition between them with gentle camera motion. --resolution 1080p"},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 11,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 API Pricing — Per-Second Rates

Resolution	Price (USD)
480p	$0.0957/sec
720p	$0.2148/sec
1080p	$0.483/sec
4k	$1.932/sec

Seedance 2.0 API Documentation

Seedance 2.0 is a high-quality AI video generator. This is the standard model — best quality, with output up to 4K.

What this does: Upload a video plus a text instruction, and it changes the video to match your request.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0/v1/edit-video

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Edit Video - Seedance 2.0 API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0/v1/edit-video
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "text", "text": "Make truck Orange"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0/v1/edit-video"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "text", "text": "Make truck Orange"},
        {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0/v1/edit-video';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'text', text: 'Make truck Orange' },
    { type: 'video_url', video_url: { url: 'https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0/v1/edit-video" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "text", "text": "Make truck Orange"},
      {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 edit-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 text + 1 image_url + 1 video_url	Edit instruction + reference object image + video to edit.
`content[].type`	Yes	string	—	"text", "image_url", "video_url"	Content item type.
`content[].text`	Yes	string	—	Recommended: under 1000 words	Edit instruction; refer to uploads by order, e.g. "Image 1", "Video 1".
`content[].image_url.url`	Yes	string	—	URL, base64 URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; max 30 MB; 300–6000 px per side	Upload an image of the object, character, or scene to bring into the edited video. JPG, PNG or WebP; under 30 MB.
`content[].video_url.url`	Yes	string	—	URL or `asset://<ID>`; mp4/mov; max 200 MB; 2–15 s; 24–60 FPS	Upload the video you want to edit. MP4; under 50 MB; 2–15 seconds.
`content[].role`	No	string	auto-assigned	"reference_image", "reference_video"	Asset role. The gateway sets it automatically from the item type; any value you send is overwritten.
`ratio`	No	string	"adaptive"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio.
`resolution`	No	string	"720p"	"480p", "720p", "1080p", "4k"	Output resolution. Higher resolutions take longer to render. 4k output is 10-bit H.265 (HEVC).
`duration`	No	integer	5	4–15, or `-1` (model auto-selects)	Output video length in seconds. Use -1 to let the model pick a suitable length.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add an "AI Generated" watermark to the bottom-right corner of the output video.
`return_last_frame`	No	boolean	false	true, false	Also return the last frame of the generated video as a PNG (same dimensions as the video, no watermark) in the status result — useful for chaining consecutive clips.
`execution_expires_after`	No	integer	172800	3600–259200	Task expiry window in seconds, counted from task creation. Tasks still queued or running past this limit are marked expired.
`priority`	No	integer	0	0–9	Queue priority. Higher values are dequeued first within your endpoint; running tasks are never interrupted.
`safety_identifier`	No	string	—	string, max 64 characters	Stable, anonymized end-user ID (for example a hash of the username) that helps the provider detect policy-violating usage.

Content Item Types

Type	Max	Format / Size	Description
video_url	1	MP4 · < 50 MB	Source video to edit (required). `{"type":"video_url","video_url":{"url":"https://..."}}`
text	1	—	Edit instructions. `{"type":"text","text":"..."}`
image_url	1	JPG, PNG, WEBP · < 30 MB	Optional style / reference image. `{"type":"image_url","image_url":{"url":"https://..."}}`

In-Prompt Directives (optional)

Seedance 2.0 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 1080p`	Output resolution (`480p` / `720p` / `1080p` / `4k`).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Restyle the source video using Image 1 color palette, framing Image 2 as the final beat. --resolution 1080p"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 11,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 API Pricing — Per-Second Rates

Resolution	Price (USD)
480p	$0.0957/sec
720p	$0.2148/sec
1080p	$0.483/sec
4k	$1.932/sec

Seedance 2.0 Fast API Documentation

Seedance 2.0 Fast is the quicker, lower-cost version of Seedance 2.0 — ideal for drafts and fast results (480p and 720p).

What this does: Describe a scene in plain words and get a short video of it.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0-fast/v1/text-to-video-fast

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Text to Video - Seedance 2.0 Fast API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0-fast/v1/text-to-video-fast
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [{"type": "text", "text": "Neon city skyline at night with flying drones"}]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0-fast/v1/text-to-video-fast"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [{"type": "text", "text": "Neon city skyline at night with flying drones"}]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0-fast/v1/text-to-video-fast';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [{ type: 'text', text: 'Neon city skyline at night with flying drones' }]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0-fast/v1/text-to-video-fast" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [{"type": "text", "text": "Neon city skyline at night with flying drones"}]
  }'

Output

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-fast",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-fast",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Request Parameters - Seedance 2.0 Fast text-to-video

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1+ text item	Input content array. Pass a single text item: `{"type":"text","text":"…"}`.
`content[].type`	Yes	string	—	"text"	Content item type.
`content[].text`	Yes	string	—	Max 500 Chinese / 1000 English words	Prompt describing the desired video.
`tools`	No	array	—	`[{"type":"web_search"}]`	Optional enhancement — text-to-video only.
`ratio`	No	string	"16:9"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio. The gateway applies "16:9" when omitted; pass "adaptive" to let the model pick the best ratio from the input.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. The Fast variant supports 480p and 720p only — 1080p is rejected by the gateway.
`duration`	No	integer	5	4–15, or `-1` for auto	Output video length in seconds. Use -1 to let the model choose the length automatically.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add a watermark to the output video.

Content Item Types

Type	Format	Description
text	`{"type":"text","text":"..."}`	The generation prompt (required). May include in-prompt directives below.

In-Prompt Directives (optional)

Seedance 2.0 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p` on the Fast variant).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Neon city skyline at night with flying drones --resolution 720p --seed 7 --framepersecond 24"}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 5,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0-fast' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-fast",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-fast",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-fast_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:01:15.000Z",
  "completed_at": "2026-04-07T10:01:15.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 Fast API Pricing

Seedance 2.0 is the production model. Mini is the cheapest tier; Fast trades quality for speed. 1.0 Pro / 1.0 Lite are legacy variants.

Your request will cost ~~$0.1111~~ $0.0833 per second.

e.g. $1.20 for a 5s video, $2.40 for a 10s video

Seedance 2.0 Fast API Documentation

Seedance 2.0 Fast is the quicker, lower-cost version of Seedance 2.0 — ideal for drafts and fast results (480p and 720p).

What this does: Give a starting image (and optionally an ending image), and it creates a video that animates between them.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0-fast/v1/first-last-frame-to-video-fast

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Image to Video (First & Last Frames) - Seedance 2.0 Fast API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0-fast/v1/first-last-frame-to-video-fast
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0-fast/v1/first-last-frame-to-video-fast"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0-fast/v1/first-last-frame-to-video-fast';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png' } },
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0-fast/v1/first-last-frame-to-video-fast" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type":"image_url","image_url":{"url":"https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
      {"type":"image_url","image_url":{"url":"https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-fast",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-fast",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 Fast image-to-video (first & last frames)

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 or 2 image_url; optional text	First-frame (1 image) or first+last-frame (2 images) mode.
`content[].type`	Yes	string	—	"text", "image_url"	Content item type.
`content[].text`	No	string	—	Max 500 Chinese / 1000 English words	Optional motion guidance prompt.
`content[].image_url.url`	Yes	string	—	URL, base64 URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; 300–6000 px; max 30 MB	Upload the image your video starts from — it becomes the first frame, then motion is added. JPG, PNG or WebP; under 30 MB.
`content[].role`	No	string	auto-assigned	"first_frame", "last_frame"	Frame position. The gateway assigns it automatically from image order (1st = first_frame, 2nd = last_frame) — you do not need to set it.
`ratio`	No	string	"16:9"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio. The gateway applies "16:9" when omitted; pass "adaptive" to let the model pick the best ratio from the input.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. The Fast variant supports 480p and 720p only — 1080p is rejected by the gateway.
`duration`	No	integer	5	4–15, or `-1` for auto	Output video length in seconds. Use -1 to let the model choose the length automatically.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add a watermark to the output video.

Content Item Types

Type	Max	Format / Size	Description
image_url	2	JPG, PNG, WEBP · < 30 MB each	1st = first frame (required); 2nd = last frame (optional). `{"type":"image_url","image_url":{"url":"https://..."}}`
text	1	—	Optional motion/style prompt. `{"type":"text","text":"..."}`

In-Prompt Directives (optional)

Seedance 2.0 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p` on the Fast variant).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Content Item Types

Type	Format	Description
`text`	`{"type":"text","text":"..."}`	Optional. Text prompt describing motion / scene transition. Maximum 1 `text` item.
`image_url` (1st)	`{"type":"image_url","image_url":{"url":"https://..."}}`	Required. First `image_url` in array order. Worker auto-injects role `first_frame`.
`image_url` (2nd)	`{"type":"image_url","image_url":{"url":"https://..."}}`	Required. Second `image_url` in array order. Worker auto-injects role `last_frame`.
`video_url`	—	Not accepted on this route (rejected with 400).
`audio_url`	—	Not accepted on this route (rejected with 400).

Image Constraints (per `image_url` item)

Constraint	Value
Format	The URL of the starting frame image to animate. Supported formats: JPEG, PNG, WebP. Max 30 MB.

Pixel dimensions per ratio

Seedance 2.0 Fast supports 480p and 720p only.

Ratio	480p	720p
`16:9`	864×496	1280×720
`4:3`	752×560	1112×834
`1:1`	640×640	960×960
`3:4`	560×752	834×1112
`9:16`	496×864	720×1280
`21:9`	992×432	1470×630

Example Request

{
  "content": [
    {"type": "text", "text": "Time-lapse transition from winter to spring: snow melts off the plant, leaves turn green, flowers bloom into full color. Smooth cinematic motion."},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "duration": 5,
  "ratio": "adaptive",
  "resolution": "720p",
  "generate_audio": true,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0-fast' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-fast",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-fast",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-fast_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 Fast API Pricing

Your request will cost ~~$0.1111~~ $0.0833 per second.

e.g. $1.20 for a 5s video, $2.40 for a 10s video

Seedance 2.0 Fast API Documentation

Seedance 2.0 Fast is the quicker, lower-cost version of Seedance 2.0 — ideal for drafts and fast results (480p and 720p).

What this does: Give it photos, clips, or audio of people or things, and it creates a video featuring them.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0-fast/v1/reference-to-video-fast

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Reference to Video - Seedance 2.0 Fast API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0-fast/v1/reference-to-video-fast
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0-fast/v1/reference-to-video-fast"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0-fast/v1/reference-to-video-fast';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'image_url', image_url: { url: 'https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0-fast/v1/reference-to-video-fast" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-fast",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-fast",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 Fast multimodal reference

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	optional text + at least one of {image_url, video_url, audio_url}; ordered text → images → videos → audio	Multimodal reference input.
`content[].type`	Yes	string	—	"text", "image_url", "video_url", "audio_url"	Content item type.
`content[].text`	No	string	—	Max 500 Chinese / 1000 English words; may include `@image1` / `@video1` / `@audio1` refs	Prompt with optional reference handles.
`content[].image_url.url`	Conditional	string	—	URL, base64 data URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; max 30 MB	Upload up to 9 reference images the model takes style or subject cues from, to keep your result consistent. JPG, PNG or WebP; under 30 MB each.
`content[].video_url.url`	Conditional	string	—	URL or `asset://<ID>`; mp4/mov; max 200 MB; 2–15 s each, ≤15 s total	Upload up to 3 reference videos the model takes motion or style cues from. MP4; under 50 MB each; up to 15 seconds total.
`content[].audio_url.url`	Conditional	string	—	URL, base64 URI, or `asset://<ID>`; wav/mp3; 2–15 s each, ≤15 s total; max 15 MB	Upload up to 3 audio clips as a reference. MP3 or WAV; under 15 MB each; up to 15 seconds total. Audio alone won't work — add at least one image or video.
`content[].role`	No	string	auto-assigned	"reference_image", "reference_video", "reference_audio"	Asset role. The gateway sets it automatically from the item type (image → reference_image, video → reference_video, audio → reference_audio); any value you send is overwritten.
`ratio`	No	string	"16:9"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio. The gateway applies "16:9" when omitted; pass "adaptive" to let the model pick the best ratio from the input.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. The Fast variant supports 480p and 720p only — 1080p is rejected by the gateway.
`duration`	No	integer	5	4–15, or `-1` for auto	Output video length in seconds. Use -1 to let the model choose the length automatically.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add a watermark to the output video.

Content Item Types & Limits

Type	Max count	Format / Size	Description
image_url	up to 9	JPG, PNG, WEBP · < 30 MB each	Reference image for character consistency, style, or composition. `{"type":"image_url","image_url":{"url":"https://..."}}`
video_url	up to 3	MP4 · < 50 MB each · ≤ 15s combined	Reference video for motion / style transfer. `{"type":"video_url","video_url":{"url":"https://..."}}`
audio_url	up to 3	MP3, WAV · ≤ 15s combined · < 15 MB	Reference audio for audio-driven generation & lip-sync. Requires at least one image or video reference. `{"type":"audio_url","audio_url":{"url":"https://..."}}`
text	1	—	Prompt. Address references by their 1-indexed array order — `[Image1]`, `[Video1]`, `[Audio1]` (plain "Image 1" also works).

The rule of 12: total references across all types may not exceed 12 (e.g. 9 images + 3 videos = 12, leaving no room for audio). Per-type caps (9 / 3 / 3) always apply within that budget.

In-Prompt Directives (optional)

Seedance 2.0 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p` on the Fast variant).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Use Image 1 as the opening frame and Image 2 as the closing frame. Smooth cinematic transition. --resolution 720p"},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 11,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0-fast' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-fast",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-fast",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-fast_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:01:15.000Z",
  "completed_at": "2026-04-07T10:01:15.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 Fast API Pricing

Your request will cost ~~$0.1111~~ $0.0833 per second.

e.g. $1.20 for a 5s video, $2.40 for a 10s video

Seedance 2.0 Fast API Documentation

Seedance 2.0 Fast is the quicker, lower-cost version of Seedance 2.0 — ideal for drafts and fast results (480p and 720p).

What this does: Upload a video plus a text instruction, and it changes the video to match your request.

What you get back: a link to a finished video file (MP4) you can download.

How it works, in 3 simple steps:

Send your request to the URL below, including your API key.
You instantly get back a request_id and a status link. Your video is then created in the background (this takes a little time).
Open the status link every few seconds until it shows COMPLETED, then download your video from the link it returns. Prefer not to keep checking? Add a webhook (see the Webhook section) and we will notify your server automatically when it is ready.

Base URL

https://gateway.pixazo.ai/seedance-2-0-fast/v1/edit-video-fast

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Edit Video - Seedance 2.0 Fast API

Request Code

POST https://gateway.pixazo.ai/seedance-2-0-fast/v1/edit-video-fast
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "content": [
    {"type": "text", "text": "Make truck Orange"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
  ]
}

import requests

url = "https://gateway.pixazo.ai/seedance-2-0-fast/v1/edit-video-fast"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "content": [
        {"type": "text", "text": "Make truck Orange"},
        {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/seedance-2-0-fast/v1/edit-video-fast';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  content: [
    { type: 'text', text: 'Make truck Orange' },
    { type: 'video_url', video_url: { url: 'https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4' } }
  ]
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -v -X POST "https://gateway.pixazo.ai/seedance-2-0-fast/v1/edit-video-fast" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "content": [
      {"type": "text", "text": "Make truck Orange"},
      {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}}
    ]
  }'

Output

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-fast",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-fast",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Using images & real people

You can supply an image or video in three ways:

A web link (URL) — a public link to your image or video file.
Base64 — the file encoded as a text string (for images and audio).
An asset ID — asset://<ID>, if you uploaded the file to the asset library earlier.

Real people are allowed. Photos and clips of real human faces work normally — the gateway automatically registers your reference media with the provider, so real-person content is not blocked. (Advanced: for formally consented real-person assets, add them in the BytePlus console and pass the resulting asset://<ID>.)

Request Parameters - Seedance 2.0 Fast video editing

Parameter	Required	Type	Default	Allowed values / range	Description
`content`	Yes	array	—	1 video_url + optional text + optional image_url	The video to edit, plus an optional edit instruction and optional reference object image.
`content[].type`	Yes	string	—	"text", "image_url", "video_url"	Content item type.
`content[].text`	No	string	—	Max 500 Chinese / 1000 English words; may use `@image1` / `@video1`	Edit instruction with reference handles (strongly recommended).
`content[].image_url.url`	No	string	—	URL, base64 URI, or `asset://<ID>`; jpeg/png/webp/bmp/tiff/gif/heic/heif; max 30 MB	Optional. Upload an image of an object, character, or scene to bring into the edited video. JPG, PNG or WebP; under 30 MB.
`content[].video_url.url`	Yes	string	—	URL or `asset://<ID>`; mp4/mov; max 200 MB; 2–15 s	Upload the video you want to edit. MP4; under 50 MB; 2–15 seconds.
`content[].role`	No	string	auto-assigned	"reference_image", "reference_video"	Asset role. The gateway sets it automatically from the item type; any value you send is overwritten.
`ratio`	No	string	"16:9"	"16:9", "4:3", "1:1", "3:4", "9:16", "21:9", "adaptive"	Output aspect ratio. The gateway applies "16:9" when omitted; pass "adaptive" to let the model pick the best ratio from the input.
`resolution`	No	string	"720p"	"480p", "720p"	Output resolution. The Fast variant supports 480p and 720p only — 1080p is rejected by the gateway.
`duration`	No	integer	5	4–15, or `-1` for auto	Output video length in seconds. Use -1 to let the model choose the length automatically.
`generate_audio`	No	boolean	true	true, false	Generate synchronized audio along with the video.
`watermark`	No	boolean	false	true, false	Add a watermark to the output video.

Content Item Types

Type	Max	Format / Size	Description
video_url	1	MP4 · < 50 MB	Source video to edit (required). `{"type":"video_url","video_url":{"url":"https://..."}}`
text	1	—	Edit instructions. `{"type":"text","text":"..."}`
image_url	1	JPG, PNG, WEBP · < 30 MB	Optional style / reference image. `{"type":"image_url","image_url":{"url":"https://..."}}`

In-Prompt Directives (optional)

Seedance 2.0 also accepts space-separated flags appended to the text prompt. If the matching JSON field is also present, the JSON field takes precedence and the directive is ignored.

Directive	Example	Description
`--resolution`	`--resolution 720p`	Output resolution (`480p` / `720p` on the Fast variant).
`--seed`	`--seed 42`	Reproducible seed.
`--framepersecond`	`--framepersecond 24`	Output frame rate (fps).
`--camerafixed`	`--camerafixed false`	`true` locks the camera; `false` allows camera motion.

Example Request

{
  "content": [
    {"type": "text", "text": "Restyle the source video using Image 1 color palette, framing Image 2 as the final beat. --resolution 720p"},
    {"type": "video_url", "video_url": {"url": "https://ark-doc.tos-ap-southeast-1.bytepluses.com/doc_video/r2v_tea_video1.mp4"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"}},
    {"type": "image_url", "image_url": {"url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f2.png"}}
  ],
  "generate_audio": true,
  "ratio": "16:9",
  "duration": 11,
  "watermark": false
}

Response

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance."
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'seedance-2-0-fast' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "seedance-2-0-fast",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "seedance-2-0-fast_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "seedance-2-0-fast",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/seedance-2-0-fast_019dxxxx-xxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-07T10:00:00.000Z",
  "updated_at": "2026-04-07T10:02:30.000Z",
  "completed_at": "2026-04-07T10:02:30.000Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type of the output
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 2.0 Fast API Pricing

Your request will cost ~~$0.1111~~ $0.0833 per second.

e.g. $1.20 for a 5s video, $2.40 for a 10s video

Seedance 1.0 Pro API Documentation

https://gateway.pixazo.ai/byteplus/v1/generateImage2VideoTask

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Seedance 1.0 Image to Video - Bytedance API

Request Code

POST https://gateway.pixazo.ai/byteplus/v1/generateImage2VideoTask
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "model": "seedance-1-0-pro-250528",
  "text": "Soft cotton-like clouds drift with subtle layered motions across a pale blue sky. --ratio 16:9 --resolution 720p --duration 5",
  "image_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"
}

import requests

url = "https://gateway.pixazo.ai/byteplus/v1/generateImage2VideoTask"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
  "model": "seedance-1-0-pro-250528",
  "text": "Soft cotton-like clouds drift with subtle layered motions across a pale blue sky. --ratio 16:9 --resolution 720p --duration 5",
  "image_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/byteplus/v1/generateImage2VideoTask';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  "model": "seedance-1-0-pro-250528",
  "text": "Soft cotton-like clouds drift with subtle layered motions across a pale blue sky. --ratio 16:9 --resolution 720p --duration 5",
  "image_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -X POST "https://gateway.pixazo.ai/byteplus/v1/generateImage2VideoTask" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
  "model": "seedance-1-0-pro-250528",
  "text": "Soft cotton-like clouds drift with subtle layered motions across a pale blue sky. --ratio 16:9 --resolution 720p --duration 5",
  "image_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"
}'

Output

{
  "request_id": "byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "bytedance-text-to-image_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "bytedance-text-to-image",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/bytedance-text-to-image_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "bytedance-text-to-image_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "bytedance-text-to-image",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Request Parameters - Seedance 1.0 Image to Video - Bytedance API

Parameter	Required	Type	Default	Allowed values / range	Description
model	Yes	string	—	seedance-1-0-pro-250528	The Seedance model ID. Use `seedance-1-0-pro-250528` for Seedance 1.0 Pro image-to-video generation.
text	Yes	string	—	—	Prompt describing the desired motion. Control knobs can be embedded inline: `--ratio`, `--resolution`, `--duration`, `--watermark`, `--seed`, `--camerafixed` (see below).
image_url	Yes	string	—	—	Publicly accessible HTTPS URL of the source image used as the first frame of the video.
resolution	No	string	720p	480p, 720p, 1080p	Output resolution of the video. Appended to the prompt as `--resolution` when not already present. Higher resolutions cost more and take longer to generate.
duration	No	integer	5	2–12	Video duration in seconds (integers only). Appended to the prompt as `--duration` when not already present.
framepersecond	No	integer	24	24	Frame rate of the generated video. Seedance 1.0 Pro renders at a fixed 24 fps.
watermark	No	boolean	false	true, false	Whether the video carries an "AI Generated" watermark in the lower-right corner.
seed	No	integer	-1	-1 to 4294967295	Seed controlling randomness. `-1` uses a random seed. The same seed with an identical request produces similar (not guaranteed identical) results. Only sent upstream when set to a value other than -1.
camerafixed	No	boolean	false	true, false	Whether to lock the camera position. When `true`, a fixed-camera instruction is appended to the prompt (result not guaranteed).
callback_url	No	string	—	—	Publicly reachable URL that receives a POST notification from the provider each time the task status changes (queued, running, succeeded, failed, expired).
Inline knobs (embedded in the `text` prompt). Each body parameter above is auto-appended as its inline knob when not already present in the prompt; if you set a knob inline, use its full name (e.g. `--resolution`, not `--rs`) so the gateway does not append a conflicting default.
--ratio	No	string	adaptive	16:9, 4:3, 1:1, 3:4, 9:16, 21:9, adaptive	Aspect ratio of the generated video. `adaptive` (default) picks the closest ratio to the uploaded image; if a fixed ratio differs from the image, the image is center-cropped. Abbreviation: `--rt`. Inline only — there is no body parameter for aspect ratio.
--resolution	No	string	720p	480p, 720p, 1080p	Inline form of `resolution`. Abbreviation: `--rs`.
--duration	No	integer	5	2–12	Inline form of `duration`. Abbreviation: `--dur`.
--watermark	No	boolean	false	true, false	Inline form of `watermark`. Abbreviation: `--wm`.
--seed	No	integer	-1	-1 to 4294967295	Inline form of `seed`.
--camerafixed	No	boolean	false	true, false	Inline form of `camerafixed`. Abbreviation: `--cf`.

Example Request

{
  "model": "seedance-1-0-pro-250528",
  "text": "Soft cotton-like clouds drift with subtle layered motions across a pale blue sky. --ratio 16:9 --resolution 720p --duration 5",
  "image_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/f1.png"
}

Response

{
  "request_id": "byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY
X-Webhook-URL	Optional callback URL

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance. Required: $0.20"
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'byteplus' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "byteplus",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "byteplus",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-09T10:00:00.000Z",
  "updated_at": "2026-04-09 10:01:30",
  "completed_at": "2026-04-09 10:01:30"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type (video/mp4)
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 1.0 Pro API Pricing

Your request will cost $0.20 per video.

Seedance 1.0 Pro API Documentation

https://gateway.pixazo.ai/byteplus/v1/generateVideoTask

Authentication

All requests require an API key passed via header.

Header	Type	Required	Description
Ocp-Apim-Subscription-Key	string	Yes	Your API subscription key

Seedance 1.0 Pro Text to Video - Bytedance API

Request Code

POST https://gateway.pixazo.ai/byteplus/v1/generateVideoTask
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "model": "seedance-1-0-pro-250528",
  "text": "A vast expanse of white daisy fields under a clear blue sky. --ratio 16:9 --resolution 720p --duration 5 --camerafixed false"
}

import requests

url = "https://gateway.pixazo.ai/byteplus/v1/generateVideoTask"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
  "model": "seedance-1-0-pro-250528",
  "text": "A vast expanse of white daisy fields under a clear blue sky. --ratio 16:9 --resolution 720p --duration 5 --camerafixed false"
}

response = requests.post(url, json=data, headers=headers)
print(response.json())

const url = 'https://gateway.pixazo.ai/byteplus/v1/generateVideoTask';
const headers = {
  'Content-Type': 'application/json',
  'Cache-Control': 'no-cache',
  'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
};
const data = {
  "model": "seedance-1-0-pro-250528",
  "text": "A vast expanse of white daisy fields under a clear blue sky. --ratio 16:9 --resolution 720p --duration 5 --camerafixed false"
};

fetch(url, {
  method: 'POST',
  headers: headers,
  body: JSON.stringify(data)
})
.then(response => response.json())
.then(data => console.log(data))
.catch(error => console.error('Error:', error));

curl -X POST "https://gateway.pixazo.ai/byteplus/v1/generateVideoTask" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
  "model": "seedance-1-0-pro-250528",
  "text": "A vast expanse of white daisy fields under a clear blue sky. --ratio 16:9 --resolution 720p --duration 5 --camerafixed false"
}'

Output

{
  "request_id": "byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Add the X-Webhook-URL header to your submit request to receive a POST callback when the job completes — no polling required.

Using curl? These are HTTP request headers — pass each with -H, e.g. -H "X-Webhook-URL: https://your-server.com/webhook/callback". Do not paste them as bare lines, and end every line of a multi-line command with \.

Webhook Headers

Header	Required	Default	Description
`X-Webhook-URL`	Yes (to enable)	—	HTTPS endpoint on your server that will receive the `POST` callback. Must respond `2xx` within a few seconds (process async if needed).
`X-Webhook-Mode`	No	`terminal`	`terminal` — fires once at the final status (`COMPLETED`/`FAILED`/`ERROR`). `sync` — fires on every poll cycle plus the terminal event, and caps the queue’s polling delay at 15s for tighter progress updates.

Example: enable webhook

X-Webhook-URL: https://your-server.com/webhook/callback
X-Webhook-Mode: terminal

Callback Payload

Your endpoint receives a POST application/json with the same shape as the GET /v2/requests/status/{request_id} response. Example terminal callback (mode terminal):

{
  "request_id": "bytedance-text-to-image_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "bytedance-text-to-image",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/bytedance-text-to-image_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-05-22T13:17:32.110Z",
  "updated_at": "2026-05-22 13:19:23",
  "completed_at": "2026-05-22 13:19:23"
}

Failure callback shape

{
  "request_id": "bytedance-text-to-image_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "bytedance-text-to-image",
  "error": "Description of the error",
  "output": null,
  "created_at": "...",
  "updated_at": "...",
  "completed_at": "..."
}

Delivery semantics

terminal mode (default) — exactly one POST when the request reaches a terminal status. No callback during PROCESSING.
sync mode — POST on every status poll (with delay capped at ~15s) plus a final POST at terminal status. Use when you want progress updates.
Idempotency — use request_id as your idempotency key. Network retries can deliver the same callback more than once; your handler must tolerate duplicates.
Response — respond 200 OK within a few seconds. The queue does not block on slow handlers, but persistent failures may stop further deliveries.
HTTPS required — plain http:// URLs are rejected.

Request Parameters - Seedance 1.0 Pro Text to Video - Bytedance API

Parameter	Required	Type	Default	Allowed values / range	Description
model	Yes	string	—	seedance-1-0-pro-250528	The Seedance model ID. Use `seedance-1-0-pro-250528` for Seedance 1.0 Pro text-to-video generation.
text	Yes	string	—	—	Prompt describing the video to generate. Control knobs can be embedded inline: `--ratio`, `--resolution`, `--duration`, `--watermark`, `--seed`, `--camerafixed` (see below).
resolution	No	string	720p	480p, 720p, 1080p	Output resolution of the video. Appended to the prompt as `--resolution` when not already present. Higher resolutions cost more and take longer to generate.
duration	No	integer	5	2–12	Video duration in seconds (integers only). Appended to the prompt as `--duration` when not already present.
framepersecond	No	integer	24	24	Frame rate of the generated video. Seedance 1.0 Pro renders at a fixed 24 fps.
watermark	No	boolean	false	true, false	Whether the video carries an "AI Generated" watermark in the lower-right corner.
seed	No	integer	-1	-1 to 4294967295	Seed controlling randomness. `-1` uses a random seed. The same seed with an identical request produces similar (not guaranteed identical) results. Only sent upstream when set to a value other than -1.
camerafixed	No	boolean	false	true, false	Whether to lock the camera position. When `true`, a fixed-camera instruction is appended to the prompt (result not guaranteed).
callback_url	No	string	—	—	Publicly reachable URL that receives a POST notification from the provider each time the task status changes (queued, running, succeeded, failed, expired).
Inline knobs (embedded in the `text` prompt). Each body parameter above is auto-appended as its inline knob when not already present in the prompt; if you set a knob inline, use its full name (e.g. `--resolution`, not `--rs`) so the gateway does not append a conflicting default.
--ratio	No	string	16:9	16:9, 4:3, 1:1, 3:4, 9:16, 21:9	Aspect ratio of the generated video. For text-to-video, `adaptive` is not supported by Seedance 1.0 Pro. Abbreviation: `--rt`. Inline only — there is no body parameter for aspect ratio.
--resolution	No	string	720p	480p, 720p, 1080p	Inline form of `resolution`. Abbreviation: `--rs`.
--duration	No	integer	5	2–12	Inline form of `duration`. Abbreviation: `--dur`.
--watermark	No	boolean	false	true, false	Inline form of `watermark`. Abbreviation: `--wm`.
--seed	No	integer	-1	-1 to 4294967295	Inline form of `seed`.
--camerafixed	No	boolean	false	true, false	Inline form of `camerafixed`. Abbreviation: `--cf`.

Example Request

{
  "model": "seedance-1-0-pro-250528",
  "text": "A vast expanse of white daisy fields under a clear blue sky. --ratio 16:9 --resolution 720p --duration 5 --camerafixed false"
}

Response

{
  "request_id": "byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	YOUR_SUBSCRIPTION_KEY
X-Webhook-URL	Optional callback URL

Response Handling

Common status codes.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
402	Insufficient Balance
403	Forbidden
429	Too Many Requests
500	Internal Server Error

Error Responses

Queue system errors and model validation errors.

Queue System Errors

// 402 — Insufficient balance
{
  "error": "Insufficient Balance",
  "message": "Your wallet does not have enough balance. Required: $0.20"
}

// 400 — Model not found
{
  "error": "Model not found",
  "message": "Model 'byteplus' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "byteplus",
  "error": "Description of the error",
  "output": null
}

Retrieving Results

Poll the universal status endpoint to check progress and retrieve results.

Endpoint

GET https://gateway.pixazo.ai/v2/requests/status/{request_id}
Ocp-Apim-Subscription-Key: YOUR_API_KEY

cURL Example

curl -H "Ocp-Apim-Subscription-Key: YOUR_API_KEY" \
  "https://gateway.pixazo.ai/v2/requests/status/byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "byteplus",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/byteplus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.mp4"
    ],
    "media_type": "video/mp4"
  },
  "created_at": "2026-04-09T10:00:00.000Z",
  "updated_at": "2026-04-09 10:01:30",
  "completed_at": "2026-04-09 10:01:30"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR
model_id	string	Model that processed the request
error	string\|null	Error message if failed
output.media_url	array	URLs to generated media (R2 CDN)
output.media_type	string	MIME type (video/mp4)
created_at	string	When request was created
completed_at	string\|null	When request completed
polling_url	string	Status URL (initial response only)

Status Values

Status	Description
QUEUED	Request accepted, waiting to be processed
PROCESSING	Being processed by the model
COMPLETED	Done — output contains the result
FAILED	Failed — check error field
ERROR	System error — not charged

Status Flow

QUEUED → PROCESSING → COMPLETED
                    → FAILED
                    → ERROR

Typical Workflow

Send a generate request to the API endpoint
Save the request_id from the response
Poll every 5-10 seconds: GET /v2/requests/status/{request_id}
When status is "COMPLETED", download from output.media_url

Tip: Use X-Webhook-URL header to get a callback instead of polling.

Seedance 1.0 Pro API Pricing

Your request will cost $0.20 per video.

Seedance 2.5 API, Seedance 2.0 Mini API, Seedance 2.0 API, 1.0 Pro & Lite API: Pricing, Documentation

Models Version

Get $5 Free Credit on First Payment

Seedance 2.5 API Documentation

Base URL

Authentication

Text to Video - Seedance 2.5 API

Request Code

Output

Webhook (Optional)

Webhook Headers

Example: enable webhook

Callback Payload

Failure callback shape

Delivery semantics

Request Parameters - Seedance 2.5 text-to-video

Content Item Types

In-Prompt Directives (optional)

Example Request

Response

Request Headers

Response Handling

Error Responses

Queue System Errors

Error via Status/Webhook

Retrieving Results

Endpoint

cURL Example

Response (Completed)

Response Fields

Status Values

Status Flow

Typical Workflow

Seedance 2.5 API Pricing — Per-Second Rates

Seedance 2.5 API Documentation

Base URL

Authentication

Image to Video (First & Last Frames) - Seedance 2.5 API

Request Code

Output

Webhook (Optional)

Webhook Headers

Example: enable webhook

Callback Payload

Failure callback shape

Delivery semantics

Using images & real people

Request Parameters - Seedance 2.5 image-to-video (first & last frames)

Content Item Types

In-Prompt Directives (optional)

Content Item Types

Image Constraints (per image_url item)

Example Request

Response

Request Headers

Response Handling

Error Responses

Queue System Errors

Error via Status/Webhook

Retrieving Results

Endpoint

cURL Example

Response (Completed)

Response Fields

Status Values

Status Flow

Typical Workflow

Seedance 2.5 API Pricing — Per-Second Rates

Seedance 2.5 API Documentation

Base URL

Authentication

Reference to Video - Seedance 2.5 API

Request Code

Output

Webhook (Optional)

Webhook Headers

Example: enable webhook

Callback Payload

Failure callback shape

Delivery semantics

Image Constraints (per `image_url` item)