Pixazo API→Models→Text to Speech API→Qwen Audio

Qwen Audio 3.0 TTS Plus API, Qwen Audio 3.0 TTS Flash API, Qwen 3.0 ASR Flash API, Qwen 3.0 TTS: Pricing, Documentation

Qwen Audio 3.0 TTS Plus API, developers can convert text into lifelike audio with multilingual support, ideal for voiceovers, virtual assistants, audiobooks, and accessibility applications.

Get API Key

Models Version

WELCOME BONUS

Get $5 Free Credit on First Payment

No strings attached — add funds and get $5 bonus instantly

Claim Your $5 →

Qwen Audio 3.0 TTS Plus API Documentation

POST https://gateway.pixazo.ai/qwen-audio-3-0-tts-plus/v1/text-to-speech

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 Speech - Qwen Audio 3.0 TTS Plus

Request Code

POST https://gateway.pixazo.ai/qwen-audio-3-0-tts-plus/v1/text-to-speech
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
  "voice": "longanlingxin",
  "format": "mp3",
  "sample_rate": 22050
}

import requests

url = "https://gateway.pixazo.ai/qwen-audio-3-0-tts-plus/v1/text-to-speech"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
    "voice": "longanlingxin",
    "format": "mp3",
    "sample_rate": 22050
}

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

const url = 'https://gateway.pixazo.ai/qwen-audio-3-0-tts-plus/v1/text-to-speech';

const data = {
  text: "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
  voice: "longanlingxin",
  format: "mp3",
  sample_rate: 22050
};

fetch(url, {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
    'Cache-Control': 'no-cache',
    'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
  },
  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/qwen-audio-3-0-tts-plus/v1/text-to-speech" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
    "voice": "longanlingxin",
    "format": "mp3",
    "sample_rate": 22050
  }'

Output

{
  "request_id": "qwen-audio-3-0-tts-plus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-audio-3-0-tts-plus_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 status-poll response shown under Retrieving Results below. Example terminal callback (mode terminal):

{
  "request_id": "qwen-audio-3-0-tts-plus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-audio-3-0-tts-plus",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/qwen-audio-3-0-tts-plus/1721597258-a1b2c3d4.mp3"
    ],
    "media_type": "audio/mpeg"
  },
  "created_at": "2026-07-22T10:00:00.000Z",
  "updated_at": "2026-07-22 10:00:08",
  "completed_at": "2026-07-22 10:00:08"
}

Failure callback shape

{
  "request_id": "qwen-audio-3-0-tts-plus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-audio-3-0-tts-plus",
  "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 - Text to Speech - Qwen Audio 3.0 TTS Plus

Field	Type	Required	Default	Allowed values / range	Description
`text`	string	Yes	—	—	The text to synthesize into speech. Must be non-empty.
`voice`	string	No	`longanlingxin`	`longanlingxin`, `longanlufeng` (see Voices table)	Voice ID for synthesis. Passing a Qwen Audio 3.0 TTS Flash voice (e.g. `longanhuan_v3.6`) will be rejected.
`format`	string	No	`mp3`	`mp3`, `wav`, `pcm`	Audio output format.
`sample_rate`	integer	No	`22050`	up to 48000	Output sample rate in Hz (up to 48000).

Billed $0.035 per 1,000 characters of input text (rounded up per 1,000-character block, minimum one block). The chosen voice, format and sample rate do not affect the price.

Voices - Qwen Audio 3.0 TTS Plus

Pass one of these as the voice parameter. Each voice supports Chinese (Mandarin) and English. Voices are not interchangeable between Qwen Audio 3.0 TTS Plus and Qwen Audio 3.0 TTS Flash.

Voice (`voice`)	Name	Languages
`longanlingxin`	Longan Lingxin — default	Chinese (Mandarin), English
`longanlufeng`	Longan Lufeng	Chinese (Mandarin), English

Minimum Request

{
  "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars."
}

Full Request (all options)

{
  "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
  "voice": "longanlingxin",
  "format": "mp3",
  "sample_rate": 22050
}

Response

{
  "request_id": "qwen-audio-3-0-tts-plus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-audio-3-0-tts-plus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	Your API subscription key

Response Handling

Common status codes for Qwen Audio 3.0 TTS Plus Text to Speech.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
403	Forbidden
404	Not Found
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 'qwen-audio-3-0-tts-plus' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "qwen-audio-3-0-tts-plus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-audio-3-0-tts-plus",
  "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/qwen-audio-3-0-tts-plus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "qwen-audio-3-0-tts-plus_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-audio-3-0-tts-plus",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/qwen-audio-3-0-tts-plus/1721597258-a1b2c3d4.mp3"
    ],
    "media_type": "audio/mpeg"
  },
  "created_at": "2026-07-22T10:00:00.000Z",
  "updated_at": "2026-07-22T10:00:08.000Z",
  "completed_at": "2026-07-22T10:00:08.000Z"
}

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 the generated MP3 audio (R2 CDN)
output.media_type	string	MIME type of the output (audio/mpeg)
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.

Qwen Audio 3.0 TTS Plus API Pricing

Your request will cost $0.035 per 1,000 characters.

2. Qwen Audio 3.0 TTS Flash

Qwen Audio 3.0 TTS Flash API Documentation

POST https://gateway.pixazo.ai/qwen-audio-3-0-tts-flash/v1/text-to-speech

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 Speech - Qwen Audio 3.0 TTS Flash

Request Code

POST https://gateway.pixazo.ai/qwen-audio-3-0-tts-flash/v1/text-to-speech
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
  "voice": "longanhuan_v3.6",
  "format": "mp3",
  "sample_rate": 22050
}

import requests

url = "https://gateway.pixazo.ai/qwen-audio-3-0-tts-flash/v1/text-to-speech"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
    "voice": "longanhuan_v3.6",
    "format": "mp3",
    "sample_rate": 22050
}

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

const url = 'https://gateway.pixazo.ai/qwen-audio-3-0-tts-flash/v1/text-to-speech';

const data = {
  text: "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
  voice: "longanhuan_v3.6",
  format: "mp3",
  sample_rate: 22050
};

fetch(url, {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
    'Cache-Control': 'no-cache',
    'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
  },
  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/qwen-audio-3-0-tts-flash/v1/text-to-speech" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
    "voice": "longanhuan_v3.6",
    "format": "mp3",
    "sample_rate": 22050
  }'

Output

{
  "request_id": "qwen-audio-3-0-tts-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-audio-3-0-tts-flash_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.

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 status-poll response shown under Retrieving Results below. Example terminal callback (mode terminal):

{
  "request_id": "qwen-audio-3-0-tts-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-audio-3-0-tts-flash",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/qwen-audio-3-0-tts-flash/1721597258-a1b2c3d4.mp3"
    ],
    "media_type": "audio/mpeg"
  },
  "created_at": "2026-07-22T10:00:00.000Z",
  "updated_at": "2026-07-22 10:00:08",
  "completed_at": "2026-07-22 10:00:08"
}

Failure callback shape

{
  "request_id": "qwen-audio-3-0-tts-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-audio-3-0-tts-flash",
  "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 - Text to Speech - Qwen Audio 3.0 TTS Flash

Field	Type	Required	Default	Allowed values / range	Description
`text`	string	Yes	—	—	The text to synthesize into speech. Must be non-empty.
`voice`	string	No	`longanhuan_v3.6`	`longanfengyue`, `longanyuanfei`, `longanlingxi`, `longanxiaoxin`, `longanhuan_v3.6`, `longjielidou_v3.6`, `longpaopao_v3.6`, `longhuohuo_v3.6`, `longchuanshu_v3.6`, `loongmary`, `loongeva_v3.6`, `loongjohn` (see Voices table)	Voice ID for synthesis. Passing a Qwen Audio 3.0 TTS Plus voice (e.g. `longanlingxin`) will be rejected.
`format`	string	No	`mp3`	`mp3`, `wav`, `pcm`	Audio output format.
`sample_rate`	integer	No	`22050`	up to 48000	Output sample rate in Hz (up to 48000).

Billed $0.03 per 1,000 characters of input text (rounded up per 1,000-character block, minimum one block). The chosen voice, format and sample rate do not affect the price.

Voices - Qwen Audio 3.0 TTS Flash

Pass one of these as the voice parameter. Each voice supports Chinese (Mandarin) and English; the loong* voices are English-focused. Voices are not interchangeable between Qwen Audio 3.0 TTS Flash and Qwen Audio 3.0 TTS Plus.

Voice (`voice`)	Name	Languages
`longanfengyue`	Longan Fengyue	Chinese (Mandarin), English
`longanyuanfei`	Longan Yuanfei	Chinese (Mandarin), English
`longanlingxi`	Longan Lingxi	Chinese (Mandarin), English
`longanxiaoxin`	Longan Xiaoxin	Chinese (Mandarin), English
`longanhuan_v3.6`	Longan Huan — default	Chinese (Mandarin), English
`longjielidou_v3.6`	Longjie Lidou	Chinese (Mandarin), English
`longpaopao_v3.6`	Long Paopao	Chinese (Mandarin), English
`longhuohuo_v3.6`	Long Huohuo	Chinese (Mandarin), English
`longchuanshu_v3.6`	Long Chuanshu	Chinese (Mandarin), English
`loongmary`	loongmary	English
`loongeva_v3.6`	loongeva	English
`loongjohn`	loongJohn	English

Minimum Request

{
  "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars."
}

Full Request (all options)

{
  "text": "Welcome to the show. Today we explore the mysteries of deep space and the stories hidden among the stars.",
  "voice": "longanhuan_v3.6",
  "format": "mp3",
  "sample_rate": 22050
}

Response

{
  "request_id": "qwen-audio-3-0-tts-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-audio-3-0-tts-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	Your API subscription key

Response Handling

Common status codes for Qwen Audio 3.0 TTS Flash Text to Speech.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
403	Forbidden
404	Not Found
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 'qwen-audio-3-0-tts-flash' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "qwen-audio-3-0-tts-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-audio-3-0-tts-flash",
  "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/qwen-audio-3-0-tts-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "qwen-audio-3-0-tts-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-audio-3-0-tts-flash",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/qwen-audio-3-0-tts-flash/1721597258-a1b2c3d4.mp3"
    ],
    "media_type": "audio/mpeg"
  },
  "created_at": "2026-07-22T10:00:00.000Z",
  "updated_at": "2026-07-22T10:00:08.000Z",
  "completed_at": "2026-07-22T10:00:08.000Z"
}

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 the generated MP3 audio (R2 CDN)
output.media_type	string	MIME type of the output (audio/mpeg)
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.

Qwen Audio 3.0 TTS Flash API Pricing

Your request will cost $0.03 per 1,000 characters.

3. Qwen 3.0 TTS

Qwen 3.0 TTS API Documentation

https://gateway.pixazo.ai/qwen3-tts-1-7b-api-401/v1/qwen3-tts-1-7b-api-request

Authentication

All requests require an API key passed via header.

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

Qwen3-TTS 1.7B API check status - Qwen3-TTS 1.7B API

Request Code

POST https://gateway.pixazo.ai/qwen3-tts-1-7b-api-401/v1/qwen3-tts-1-7b-api-request-result
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "text": "Welcome to Pixazo. This speech was generated with Qwen3-TTS.",
  "voice": "Ryan",
  "language": "English"
}

import requests

url = "https://gateway.pixazo.ai/qwen3-tts-1-7b-api-401/v1/qwen3-tts-1-7b-api-request-result"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "text": "Welcome to Pixazo. This speech was generated with Qwen3-TTS.",
    "voice": "Ryan",
    "language": "English"
}

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

const url = 'https://gateway.pixazo.ai/qwen3-tts-1-7b-api-401/v1/qwen3-tts-1-7b-api-request-result';

const data = {
  text: 'Welcome to Pixazo. This speech was generated with Qwen3-TTS.',
  voice: 'Ryan',
  language: 'English'
};

fetch(url, {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
    'Cache-Control': 'no-cache',
    'Ocp-Apim-Subscription-Key': 'YOUR_SUBSCRIPTION_KEY'
  },
  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/qwen3-tts-1-7b-api-401/v1/qwen3-tts-1-7b-api-request-result" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "text": "Welcome to Pixazo. This speech was generated with Qwen3-TTS.",
    "voice": "Ryan",
    "language": "English"
  }'

Output

{
  "request_id": "qwen3-tts-1-7b-api-401_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen3-tts-1-7b-api-401_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.

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": "qwen3-tts-1-7b-api-401_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen3-tts-1-7b-api-401",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen3-tts-1-7b-api-401_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.wav"
    ],
    "media_type": "audio/wav"
  },
  "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": "qwen3-tts-1-7b-api-401_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen3-tts-1-7b-api-401",
  "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 - Qwen3-TTS 1.7B API check status

Parameter	Required	Type	Default	Allowed values / range	Description
text	Yes	string	—	—	The text to be converted to speech. Longer text takes proportionally longer to synthesize.
voice	No	string	Ryan	Vivian, Serena, Uncle_Fu, Dylan, Eric, Ryan, Aiden, Ono_Anna, Sohee	Pre-trained voice used for synthesis; each voice is tuned for a primary language. Ignored when `speaker_voice_embedding_file_url` is supplied. If you send neither a voice nor a speaker embedding, the gateway applies `Ryan`.
language	No	string	Auto	Auto, English, Chinese, Spanish, French, German, Italian, Japanese, Korean, Portuguese, Russian	Language of the spoken output. `Auto` detects the language from `text`.
prompt	No	string	—	—	Optional style instruction for the delivery, e.g. "Very happy." Ignored when a speaker embedding is supplied.
speaker_voice_embedding_file_url	No	string (URL)	—	Public HTTPS URL to a `.safetensors` file	URL of a speaker-embedding file produced by the Qwen 3 TTS clone-voice endpoint. When set, the cloned voice is used instead of `voice`.
reference_text	No	string	—	—	The text that was spoken in the reference audio used to create the speaker embedding. Providing it improves quality when synthesizing with a cloned voice.
temperature	No	number	0.9	0 to 1	Sampling temperature. Higher values make the delivery more varied; lower values make it more deterministic.
top_k	No	integer	50	0 or greater	Top-k sampling: how many of the highest-probability tokens are considered at each step.
top_p	No	number	1	0 to 1	Nucleus sampling threshold; only tokens within this cumulative probability are considered.
repetition_penalty	No	number	1.05	0 or greater	Penalty applied to repeated tokens/codes. Raise it to reduce stuttering or repeated sounds.
max_new_tokens	No	integer	200	1 to 8192	Maximum number of new codec tokens to generate; caps the length of the generated audio.
subtalker_dosample	No	boolean	true	true, false	Enables sampling for the sub-talker stage. Set to `false` for greedy, more deterministic decoding.
subtalker_temperature	No	number	0.9	0 to 1	Sampling temperature for the sub-talker stage.
subtalker_top_k	No	integer	50	0 or greater	Top-k sampling parameter for the sub-talker stage.
subtalker_top_p	No	number	1	0 to 1	Top-p (nucleus) sampling parameter for the sub-talker stage.

Example Request

{
  "text": "Welcome to Pixazo. This speech was generated with Qwen3-TTS.",
  "voice": "Ryan",
  "language": "English"
}

Response

{
  "request_id": "qwen3-tts-1-7b-api-401_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen3-tts-1-7b-api-401_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 'qwen3-tts-1-7b-api-401' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "qwen3-tts-1-7b-api-401_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen3-tts-1-7b-api-401",
  "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/qwen3-tts-1-7b-api-401_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "qwen3-tts-1-7b-api-401_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen3-tts-1-7b-api-401",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen3-tts-1-7b-api-401_019dxxxx-xxxx/output.ext"
    ],
    "media_type": "application/octet-stream"
  },
  "created_at": "2026-03-31T10:00:00.000Z",
  "updated_at": "2026-03-31T10:00:15.000Z",
  "completed_at": "2026-03-31T10:00:15.000Z"
}

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.

Qwen 3.0 TTS API Pricing

Your request will cost $0.09 per 1,000 characters.

about $0.11 for a typical 200-word paragraph

equivalent to $90 per 1M characters

4. Qwen 3.0 ASR Flash

Qwen 3.0 ASR Flash API Documentation

Transcribe audio into text with automatic language detection across 11 languages and Chinese dialects, plus speaker emotion detection. Asynchronous: submit returns a request_id; poll the status endpoint until the request is COMPLETED. The transcript is returned as a JSON file.

POST https://gateway.pixazo.ai/qwen3-asr-flash/v1/speech-to-text

Authentication

All requests require an API key passed via header.

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

Speech to Text - Qwen 3.0 ASR Flash

Request Code

POST https://gateway.pixazo.ai/qwen3-asr-flash/v1/speech-to-text
Content-Type: application/json
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "language": "en",
  "context": "Acme Corp, Dr. Yamada, quarterly earnings"
}

import requests

url = "https://gateway.pixazo.ai/qwen3-asr-flash/v1/speech-to-text"
headers = {
    "Content-Type": "application/json",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "language": "en",
  "context": "Acme Corp, Dr. Yamada, quarterly earnings"
}

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

const res = await fetch("https://gateway.pixazo.ai/qwen3-asr-flash/v1/speech-to-text", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
  },
  body: JSON.stringify({
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "language": "en",
  "context": "Acme Corp, Dr. Yamada, quarterly earnings"
})
});
console.log(await res.json());

curl -X POST 'https://gateway.pixazo.ai/qwen3-asr-flash/v1/speech-to-text' \
  -H 'Content-Type: application/json' \
  -H 'Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY' \
  --data-raw '{"audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3", "language": "en", "context": "Acme Corp, Dr. Yamada, quarterly earnings"}'

Output

{
  "request_id": "qwen3-asr-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen3-asr-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Instead of polling, you can receive a Webhook callback when the request reaches a terminal state. Provide a Webhook URL via header on the submit request.

Header	Required	Description
X-Webhook-URL	To enable	HTTPS URL to receive the Webhook callback.
X-Webhook-Mode	No	`terminal` (default, one callback on COMPLETED/ERROR) or `sync` (per-poll callbacks).

Example: enable Webhook

curl -X POST 'https://gateway.pixazo.ai/qwen3-asr-flash/v1/speech-to-text' \
  -H 'Content-Type: application/json' \
  -H 'Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY' \
  -H 'X-Webhook-URL: https://your-server.com/webhook' \
  --data-raw '{"audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3", "language": "en", "context": "Acme Corp, Dr. Yamada, quarterly earnings"}'

Callback Payload (success)

{
  "request_id": "qwen3-asr-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen3-asr-flash",
  "output": {
    "media_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/{request_id}/output.json",
    "media_type": "application/json"
  },
  "created_at": "2026-07-31T11:34:16.102Z",
  "completed_at": "2026-07-31T11:34:20.123Z"
}

Failure callback shape

{
  "request_id": "qwen3-asr-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen3-asr-flash",
  "error": "Description of the failure"
}

Delivery semantics

terminal mode: one Webhook callback when the request is COMPLETED or ERROR.
sync mode: a Webhook callback on each status change.
Callbacks are idempotent on request_id — de-duplicate on it.
Respond 200 within a few seconds; the Webhook endpoint must be HTTPS.

Request Parameters

Parameter	Required	Type	Default	Allowed values / range	Description
`audio_url`	Yes	string	—	http(s) URL, file up to 10 MB	Public URL of the audio to transcribe. The file is fetched by the gateway, so the URL must be reachable without authentication.
`language`	No	string	— (auto-detected)	ISO language code, e.g. `en`, `zh`, `ja`	Optional language hint. Omit it and the model detects the language automatically — the detected language is returned in the transcript either way.
`context`	No	string	—	free text	Names, product terms, acronyms or jargon that appear in the audio. Biases recognition towards this vocabulary, which improves accuracy on proper nouns and domain terminology.
`format`	No	string	from file extension	`wav`, `mp3`, `pcm`, `opus`, `speex`, `aac`, `amr`, `m4a`, `flac`, `ogg`, `webm`	Container format of the audio. Derived from the `audio_url` file extension; supply it explicitly when the URL has no recognisable extension.
`enable_itn`	No	boolean	`false`	`true`, `false`	Inverse text normalization — render spoken numbers, dates and amounts as digits. Chinese and English only.
`sample_rate`	No	integer	—	e.g. `16000`	Sample rate in Hz. Only needed for raw `pcm` input, which carries no header.

Audio limits

The audio file must be at most 10 MB. Larger files are rejected before transcription starts.
audio_url must be an http or https URL and must be publicly reachable.
Cost scales with the length of the audio (see Pricing).

Transcript Format

The completed request returns output.media_url, a JSON file containing the transcript. Fetch that URL to read the result.

{
  "text": "Senior staff, Principal Doris Jackson, Wakefield faculty, and of course my fellow classmates. I am honored to have been chosen to speak today.",
  "language": "en",
  "emotion": "happy",
  "duration": 17
}

Field	Type	Description
text	string	The full transcript, with punctuation and capitalisation.
language	string	Detected language code, e.g. `en` or `zh`. Returned whether or not you supplied a hint.
emotion	string	Detected speaker emotion, e.g. `happy` or `neutral`.
duration	number	Length of the transcribed audio in seconds. This is the quantity you are billed on.

This model returns a whole-transcript result. Word-level timestamps are not available.

Example Request

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "language": "en",
  "context": "Acme Corp, Dr. Yamada, quarterly earnings"
}

Example Response

{
  "request_id": "qwen3-asr-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen3-asr-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Required	Description
Content-Type	Yes	`application/json`
Ocp-Apim-Subscription-Key	Yes	Your API subscription key.
X-Webhook-URL	No	Enable Webhook callbacks (see Webhook section).

Response Handling

Status Code	Meaning
202	Accepted — request queued; returns `request_id` and `polling_url`.
400	Bad request — missing `audio_url`, a non-http(s) URL, or an audio format that could not be determined from the URL and was not supplied.
401	Unauthorized — missing or invalid subscription key.
402	Insufficient balance.
429	Too many requests.
500	Internal server error.

An unreachable audio_url, an oversized file, or audio containing no detectable speech is reported through the status endpoint as status: "ERROR", not as a synchronous error. Failed requests are not billed.

Retrieving Results

Poll the status endpoint with the request_id from the submit response until status is COMPLETED (or FAILED/ERROR), then fetch output.media_url for the transcript.

curl 'https://gateway.pixazo.ai/v2/requests/status/qwen3-asr-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx' \
  -H 'Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY'

Completed response

{
  "request_id": "qwen3-asr-flash_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen3-asr-flash",
  "output": {
    "media_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/{request_id}/output.json",
    "media_type": "application/json"
  },
  "created_at": "2026-07-31T11:34:16.102Z",
  "completed_at": "2026-07-31T11:34:20.123Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier.
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR.
model_id	string	The model that handled the request.
output.media_url	string	URL of the transcript JSON file.
output.media_type	string	`application/json`.
created_at	string	Request creation timestamp.
completed_at	string	Completion timestamp.
error	string	Error message when `status` is FAILED/ERROR.

Status Values & Flow

QUEUED → PROCESSING → COMPLETED (success) or FAILED/ERROR (failure).

Pricing

Billed at $0.00282 per minute of audio ($0.1692 per audio hour), rounded up to the next whole minute. Billing is based on the length of the audio you submit, not on the size of the transcript.

Audio length	Billed minutes	Cost
20 seconds	1	$0.00282
90 seconds	2	$0.00564
10 minutes	10	$0.0282
1 hour	60	$0.1692

A hold is placed when the request is submitted, because the audio length is not known until transcription completes. The hold is reduced to the real cost once the duration is known, and released in full if the request fails.

Qwen 3.0 ASR Flash API Pricing

Your request will cost $0.0028 per minute of audio, rounded up.

about $0.03 for a 10-minute recording; billed on the audio you submit, not the transcript

equivalent to $0.1692 per audio hour

5. Qwen 3.0 ASR Flash Filetrans

Qwen 3.0 ASR Flash Filetrans API Documentation

Transcribe long audio and video — up to 2 GB and 12 hours per file — with automatic language detection, per-sentence emotion and sentence-level timestamps. Asynchronous: submit returns a request_id; poll the status endpoint until the request is COMPLETED. The transcript is returned as a JSON file.

POST https://gateway.pixazo.ai/qwen3-asr-flash-filetrans/v1/speech-to-text

Authentication

All requests require an API key passed via header.

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

Speech to Text - Qwen 3.0 ASR Flash Filetrans

Request Code

POST https://gateway.pixazo.ai/qwen3-asr-flash-filetrans/v1/speech-to-text
Content-Type: application/json
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "duration": 3600
}

import requests

url = "https://gateway.pixazo.ai/qwen3-asr-flash-filetrans/v1/speech-to-text"
headers = {
    "Content-Type": "application/json",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "duration": 3600
}

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

const res = await fetch("https://gateway.pixazo.ai/qwen3-asr-flash-filetrans/v1/speech-to-text", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
  },
  body: JSON.stringify({
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "duration": 3600
})
});
console.log(await res.json());

curl -X POST 'https://gateway.pixazo.ai/qwen3-asr-flash-filetrans/v1/speech-to-text' \
  -H 'Content-Type: application/json' \
  -H 'Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY' \
  --data-raw '{"audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3", "duration": 3600}'

Output

{
  "request_id": "qwen3-asr-flash-filetrans_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen3-asr-flash-filetrans_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Try Now

Webhook (Optional)

Instead of polling, you can receive a Webhook callback when the request reaches a terminal state. Provide a Webhook URL via header on the submit request.

Header	Required	Description
X-Webhook-URL	To enable	HTTPS URL to receive the Webhook callback.
X-Webhook-Mode	No	`terminal` (default, one callback on COMPLETED/ERROR) or `sync` (per-poll callbacks).

Example: enable Webhook

curl -X POST 'https://gateway.pixazo.ai/qwen3-asr-flash-filetrans/v1/speech-to-text' \
  -H 'Content-Type: application/json' \
  -H 'Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY' \
  -H 'X-Webhook-URL: https://your-server.com/webhook' \
  --data-raw '{"audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3", "duration": 3600}'

Callback Payload (success)

{
  "request_id": "qwen3-asr-flash-filetrans_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen3-asr-flash-filetrans",
  "output": {
    "media_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/{request_id}/output.json",
    "media_type": "application/json"
  },
  "created_at": "2026-07-31T11:34:16.102Z",
  "completed_at": "2026-07-31T11:34:20.123Z"
}

Failure callback shape

{
  "request_id": "qwen3-asr-flash-filetrans_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen3-asr-flash-filetrans",
  "error": "Description of the failure"
}

Delivery semantics

terminal mode: one Webhook callback when the request is COMPLETED or ERROR.
sync mode: a Webhook callback on each status change.
Callbacks are idempotent on request_id — de-duplicate on it.
Respond 200 within a few seconds; the Webhook endpoint must be HTTPS.

Request Parameters

Parameter	Required	Type	Default	Allowed values / range	Description
`audio_url`	Yes	string	—	http(s) URL, file up to 2 GB / 12 hours	Public URL of the audio or video file to transcribe. The file is fetched by the speech service directly, so the URL must remain publicly reachable, without authentication, until the job finishes.
`duration`	No	integer	12 hours	1 – 43200 (seconds)	Length of the audio in seconds. Optional, and it never changes what you are charged — you are always billed on the speech the model actually finds. Supplying it reserves only what the job will cost instead of the 12-hour maximum, which matters when your balance is low.

Audio limits

Up to 2 GB and 12 hours per file.
Audio and video are both accepted: mp3, wav, m4a, flac, ogg, opus, aac, amr, wma, mp4, mkv, mov, avi, flv, webm, wmv, mpeg. Audio is extracted from video automatically.
audio_url must be an http or https URL. Unlike the short-form endpoint, the file is not copied to our storage first — it is read straight from your URL, which must stay reachable for the whole job.
Cost scales with the length of the speech in the audio (see Pricing).

Transcript Format

The completed request returns output.media_url, a JSON file containing the transcript. Fetch that URL to read the result.

{
  "text": "Merry Christmas to all, and to all a good night!",
  "language": "en",
  "emotion": "happy",
  "duration": 6,
  "audio_info": {
    "format": "mp3",
    "sample_rate": 44100
  },
  "sentences": [
    {
      "sentence_id": 0,
      "begin_time": 0,
      "end_time": 6250,
      "language": "en",
      "emotion": "happy",
      "text": "Merry Christmas to all, and to all a good night!"
    }
  ]
}

Field	Type	Description
text	string	The full transcript, with punctuation and capitalisation.
language	string	Detected language code, e.g. `en` or `zh`. Returned whether or not you supplied a hint.
emotion	string	Detected speaker emotion, e.g. `happy` or `neutral`.
duration	number	Seconds of speech found in the audio, excluding silence. This is the quantity you are billed on.
audio_info	object	Detected container `format` and `sample_rate` of the source file.
sentences	array	Sentence-level segments: `{sentence_id, begin_time, end_time, language, emotion, text}`, with times in milliseconds.

Sentences carry their own language and emotion, so mixed-language recordings are labelled per sentence rather than per file.

Example Request

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "duration": 3600
}

Example Response

{
  "request_id": "qwen3-asr-flash-filetrans_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen3-asr-flash-filetrans_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Required	Description
Content-Type	Yes	`application/json`
Ocp-Apim-Subscription-Key	Yes	Your API subscription key.
X-Webhook-URL	No	Enable Webhook callbacks (see Webhook section).

Response Handling

Status Code	Meaning
202	Accepted — request queued; returns `request_id` and `polling_url`.
400	Bad request — missing `audio_url`, a non-http(s) URL, or a file larger than 2 GB.
401	Unauthorized — missing or invalid subscription key.
402	Insufficient balance.
429	Too many requests.
500	Internal server error.

Retrieving Results

Poll the status endpoint with the request_id from the submit response until status is COMPLETED (or FAILED/ERROR), then fetch output.media_url for the transcript.

curl 'https://gateway.pixazo.ai/v2/requests/status/qwen3-asr-flash-filetrans_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx' \
  -H 'Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY'

Completed response

{
  "request_id": "qwen3-asr-flash-filetrans_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen3-asr-flash-filetrans",
  "output": {
    "media_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/{request_id}/output.json",
    "media_type": "application/json"
  },
  "created_at": "2026-07-31T11:34:16.102Z",
  "completed_at": "2026-07-31T11:34:20.123Z"
}

Response Fields

Field	Type	Description
request_id	string	Unique request identifier.
status	string	QUEUED, PROCESSING, COMPLETED, FAILED, or ERROR.
model_id	string	The model that handled the request.
output.media_url	string	URL of the transcript JSON file.
output.media_type	string	`application/json`.
created_at	string	Request creation timestamp.
completed_at	string	Completion timestamp.
error	string	Error message when `status` is FAILED/ERROR.

Status Values & Flow

QUEUED → PROCESSING → COMPLETED (success) or FAILED/ERROR (failure).

Pricing

Billed at $0.00282 per minute of speech ($0.1692 per hour), rounded up to the next whole minute. Billing counts only the speech in the file — silence, music and dead air are excluded, so a sparse recording costs less than its wall-clock length.

Audio length	Billed minutes	Cost
20 seconds of speech	1	$0.00282
10 minutes of speech	10	$0.0282
1 hour of speech	60	$0.1692
12 hours of speech	720	$2.0304

A hold is placed when the request is submitted, because the amount of speech is not known until transcription completes. Without a duration hint the hold is the 12-hour maximum ($2.0304); supplying duration reserves only what the job will cost. Either way the hold is reduced to the real cost on completion, and released in full if the request fails.

Qwen 3.0 ASR Flash Filetrans API Pricing

Your request will cost $0.0028 per minute of speech, rounded up.

about $0.17 for a one-hour podcast; silence and music are not billed

equivalent to $0.1692 per hour of speech

6. Qwen 3.0 ASR Flash Realtime

Qwen 3.0 ASR Flash Realtime API Documentation

Live speech-to-text over a WebSocket. Stream audio as it is captured and receive transcripts as the words arrive — partial results while someone is still speaking, then a final result per utterance with detected language and emotion. Built for live captions, voice agents and dictation.

WSS wss://asr-stream.pixazo.ai/v1/stream

This is a streaming API, so unlike our other speech-to-text models there is no submit-and-poll cycle and no request_id. The connection is the request.

Authentication

Pass your subscription key either as a query parameter or as a header. The query form exists because browsers cannot set headers on a WebSocket connection.

wss://asr-stream.pixazo.ai/v1/stream?key=YOUR_SUBSCRIPTION_KEY&language=en

Method	Where	Use when
`?key=`	query string	Browser clients
Ocp-Apim-Subscription-Key	header	Server-side clients

Treat a key in a URL with the same care as any other credential — it can appear in proxy and browser logs. For public-facing pages, proxy the connection through your own backend rather than shipping the key to the browser.

Speech to Text (Streaming) - Qwen 3.0 ASR Flash Realtime

Client Code

const ws = new WebSocket(
  "wss://asr-stream.pixazo.ai/v1/stream?key=" + encodeURIComponent(PIXAZO_API_KEY)
);

ws.onopen = () => {
  ws.send(JSON.stringify({
    type: "session.update",
    session: {
      modalities: ["text"],
      input_audio_format: "pcm",
      sample_rate: 16000,
      turn_detection: { type: "server_vad" }
    }
  }));
};

// Send 16-bit mono PCM at 16 kHz, about 100 ms (3200 bytes) per frame.
function sendAudio(pcmChunk) {
  ws.send(JSON.stringify({
    type: "input_audio_buffer.append",
    audio: btoa(String.fromCharCode(...new Uint8Array(pcmChunk)))
  }));
}

ws.onmessage = (ev) => {
  const msg = JSON.parse(ev.data);
  if (msg.type.endsWith("transcription.text")) {
    console.log("partial:", msg.text);          // updates as they speak
  }
  if (msg.type.endsWith("transcription.completed")) {
    console.log("final:", msg.transcript, msg.language, msg.emotion);
  }
};

import base64, json, websockets, asyncio

URL = "wss://asr-stream.pixazo.ai/v1/stream?key=" + PIXAZO_API_KEY

async def transcribe(pcm_stream):
    async with websockets.connect(URL) as ws:
        await ws.send(json.dumps({
            "type": "session.update",
            "session": {
                "modalities": ["text"],
                "input_audio_format": "pcm",
                "sample_rate": 16000,
                "turn_detection": {"type": "server_vad"},
            },
        }))

        async def send():
            # 3200 bytes = 100 ms of 16-bit mono PCM at 16 kHz
            for chunk in pcm_stream:
                await ws.send(json.dumps({
                    "type": "input_audio_buffer.append",
                    "audio": base64.b64encode(chunk).decode(),
                }))
                await asyncio.sleep(0.1)
            await ws.send(json.dumps({"type": "input_audio_buffer.commit"}))

        asyncio.create_task(send())
        async for raw in ws:
            msg = json.loads(raw)
            if msg["type"].endswith("transcription.completed"):
                print(msg["transcript"], msg["language"], msg["emotion"])

Final Result Event

{
  "type": "conversation.item.input_audio_transcription.completed",
  "item_id": "item_MaUP928DS8dtUSntYnfcK",
  "transcript": "Merry Christmas to all, and to all a good night!",
  "language": "en",
  "emotion": "happy"
}

Session Protocol

After connecting, configure the session, then stream audio frames. Results arrive continuously.

1. Configure the session

{
  "type": "session.update",
  "session": {
    "modalities": [
      "text"
    ],
    "input_audio_format": "pcm",
    "sample_rate": 16000,
    "input_audio_transcription": {
      "language": "en"
    },
    "turn_detection": {
      "type": "server_vad"
    }
  }
}

2. Stream audio frames

{
  "type": "input_audio_buffer.append",
  "audio": "<base64 pcm frame>"
}

With server_vad turn detection the service segments speech for you and emits one final result per utterance. Send {"type":"input_audio_buffer.commit"} to flush the last utterance before closing.

Request Parameters

Parameter	Required	Type	Default	Allowed values / range	Description
`key`	Yes	string	—	your subscription key	Query-string form of the API key. Browsers cannot set headers on a WebSocket, so the key goes in the URL. Server-side clients may instead send it as the `Ocp-Apim-Subscription-Key` or `X-Subscription-Key` header.
`language`	No	string	— (auto-detected)	ISO code, e.g. `en`, `zh`	Optional language hint for the session. Omit it and the model detects the language automatically — the detected language is returned on every final result either way.

Audio format

16-bit signed PCM, mono, 16 kHz, little-endian, base64 encoded.
About 100 ms per frame — 3200 bytes. Larger frames add latency; much smaller ones add overhead.
Send audio at roughly real time. Sending far faster than real time gives no benefit.

Server Events

Event	Meaning
session.created / session.updated	Session is ready and your configuration was applied.
input_audio_buffer.speech_started	Speech detected — an utterance has begun.
input_audio_buffer.speech_stopped	The speaker paused; the utterance is being finalised.
conversation.item.input_audio_transcription.text	Partial. The transcript so far for the current utterance; it will be revised.
conversation.item.input_audio_transcription.completed	Final. The settled transcript for one utterance, with `language` and `emotion`.
error	Something went wrong; the message explains what.

Partial result

{
  "type": "conversation.item.input_audio_transcription.text",
  "item_id": "item_MaUP928DS8dtUSntYnfcK",
  "text": "Merry Christmas to all, and"
}

Render partials as provisional text and replace them when the matching completed event arrives — partials are revised as more audio is heard.

Session Limits

Limit	Value	What happens
Maximum session length	15 minutes	The connection closes with reason `max_session_reached`. Open a new one to continue.
Idle timeout	60 seconds	A session with no audio closes with reason `idle_timeout` and is not charged.
Concurrent sessions	Fair use	Each open session reserves funds for its maximum length until it closes.

For continuous transcription beyond 15 minutes, open a fresh connection when the previous one closes. For pre-recorded files, use the file-based models instead — they accept up to 12 hours in one request and need no streaming client.

Connection Errors

Authentication and billing are resolved during the WebSocket handshake, so failures appear as an HTTP status on the upgrade request, before any session exists.

Status	Meaning
101	Connected — the session is live.
401	Missing or invalid subscription key.
402	Insufficient balance to reserve a session.
426	You sent a plain HTTP request; this endpoint requires a WebSocket upgrade.
503	The service is temporarily unable to verify your key or reach the speech backend.

Once connected, problems arrive as an error event on the socket rather than as a status code.

Pricing

Billed at $0.00282 per minute of audio streamed ($0.1692 per hour), rounded up to the next whole minute. You are charged for the audio you send, measured on our side — silent air you stream still counts, so stop sending when there is nothing to transcribe.

Audio streamed	Billed minutes	Cost
30 seconds	1	$0.00282
5 minutes	5	$0.0141
15 minutes (one full session)	15	$0.0423

Opening a session reserves the cost of a full 15 minutes. The reservation is reduced to what you actually streamed when the session closes, and released entirely if no audio was sent.

Qwen 3.0 ASR Flash Realtime API Pricing

Your request will cost $0.0028 per minute of audio streamed, rounded up.

about $0.04 for a full 15-minute session

equivalent to $0.1692 per hour streamed

7. Qwen 3 TTS 0.6B

Qwen 3 TTS 0.6B API Documentation

https://gateway.pixazo.ai/qwen-3-tts-text-to-speech-0-6b/v1/qwen-3-tts-text-to-speech-0-6b-request

Authentication

All requests require an API key passed via header.

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

Qwen 3 TTS - Text to Speech [0.6B] generate request - Qwen 3 TTS - Text to Speech [0.6B]

Request Code

POST https://gateway.pixazo.ai/qwen-3-tts-text-to-speech-0-6b/v1/qwen-3-tts-text-to-speech-0-6b-request
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "text": "The morning sun cast golden rays across the quiet village, as the sound of birdsong filled the crisp autumn air.",
  "prompt": "Very happy.",
  "voice": "Vivian",
  "language": "English"
}

import requests

url = "https://gateway.pixazo.ai/qwen-3-tts-text-to-speech-0-6b/v1/qwen-3-tts-text-to-speech-0-6b-request"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "text": "The morning sun cast golden rays across the quiet village, as the sound of birdsong filled the crisp autumn air.",
    "prompt": "Very happy.",
    "voice": "Vivian",
    "language": "English"
}

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

const url = "https://gateway.pixazo.ai/qwen-3-tts-text-to-speech-0-6b/v1/qwen-3-tts-text-to-speech-0-6b-request";

const headers = {
  "Content-Type": "application/json",
  "Cache-Control": "no-cache",
  "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
};

const data = {
  "text": "The morning sun cast golden rays across the quiet village, as the sound of birdsong filled the crisp autumn air.",
  "prompt": "Very happy.",
  "voice": "Vivian",
  "language": "English"
};

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/qwen-3-tts-text-to-speech-0-6b/v1/qwen-3-tts-text-to-speech-0-6b-request" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "text": "The morning sun cast golden rays across the quiet village, as the sound of birdsong filled the crisp autumn air.",
    "prompt": "Very happy.",
    "voice": "Vivian",
    "language": "English"
  }'

Output

{
  "request_id": "qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-3-tts-text-to-speech-0-6b_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.

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": "qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-3-tts-text-to-speech-0-6b",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.wav"
    ],
    "media_type": "audio/wav"
  },
  "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": "qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-3-tts-text-to-speech-0-6b",
  "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 - Qwen 3 TTS - Text to Speech [0.6B] generate request

Field	Type	Required	Default	Description
text	string	Yes	—	The input text to be converted into speech.
prompt	string	Yes	—	Emotional or stylistic prompt to guide voice expression (e.g., "Very happy.", "Sad and slow.").
voice	string	Yes	—	Name of the pre-trained voice to use (e.g., "Vivian"). For custom voices, use the assigned voice ID.
language	string	Yes	—	Language of the input text (e.g., "English", "Spanish", "Chinese").
top_k	integer	No	50	Controls diversity of token selection during generation. Higher values increase randomness.
top_p	float	No	1.0	Nucleus sampling threshold. Lower values focus on high-probability tokens.
temperature	float	No	0.9	Controls randomness in token sampling. Higher values increase creativity.
repetition_penalty	float	No	1.05	Penalizes repeated tokens to reduce redundancy.
subtalker_dosample	boolean	No	true	Enables sampling for subtalker model. Disable for deterministic output.
subtalker_top_k	integer	No	50	Top-k sampling parameter for subtalker model.
subtalker_top_p	float	No	1.0	Nucleus sampling threshold for subtalker model.
subtalker_temperature	float	No	0.9	Temperature parameter for subtalker model.
max_new_tokens	integer	No	200	Maximum number of tokens to generate in the output audio sequence.

Minimum Request

{
  "text": "Hello world.",
  "prompt": "Normal",
  "voice": "Vivian",
  "language": "English"
}

Full Request (all options)

{
  "text": "The morning sun cast golden rays across the quiet village, as the sound of birdsong filled the crisp autumn air.",
  "prompt": "Very happy.",
  "voice": "Vivian",
  "language": "English",
  "top_k": 50,
  "top_p": 1.0,
  "temperature": 0.9,
  "repetition_penalty": 1.05,
  "subtalker_dosample": true,
  "subtalker_top_k": 50,
  "subtalker_top_p": 1.0,
  "subtalker_temperature": 0.9,
  "max_new_tokens": 200
}

Response

{
  "request_id": "qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	Your API subscription key

Response Handling

Common status codes for Qwen 3 TTS - Text to Speech [0.6B] generate request.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
403	Forbidden
404	Not Found
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 'qwen-3-tts-text-to-speech-0-6b' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-3-tts-text-to-speech-0-6b",
  "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/qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-3-tts-text-to-speech-0-6b",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen-3-tts-text-to-speech-0-6b_019dxxxx-xxxx/output.ext"
    ],
    "media_type": "application/octet-stream"
  },
  "created_at": "2026-03-31T10:00:00.000Z",
  "updated_at": "2026-03-31T10:00:15.000Z",
  "completed_at": "2026-03-31T10:00:15.000Z"
}

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.

Qwen 3 TTS 0.6B API Pricing

Your request will cost $0.07 per 1,000 characters.

the lighter, faster Qwen 3 TTS tier at $0.07 per 1,000 characters

equivalent to $70 per 1M characters

8. Qwen 3 TTS Voice Design

Qwen 3 TTS Voice Design API Documentation

https://gateway.pixazo.ai/qwen-3-tts-voice-design-1-7b/v1/qwen-3-tts-voice-design-1-7b-request

Authentication

All requests require an API key passed via header.

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

Qwen 3 TTS Voice Design [1.7B] generate request - Qwen 3 TTS Voice Design [1.7B]

Request Code

POST https://gateway.pixazo.ai/qwen-3-tts-voice-design-1-7b/v1/qwen-3-tts-voice-design-1-7b-request
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "text": "The old clock tower chimed midnight as the fog rolled in from the harbor, wrapping the cobblestone streets in a blanket of grey.",
  "prompt": "Speak in a mysterious and atmospheric tone.",
  "language": "English"
}

import requests

url = "https://gateway.pixazo.ai/qwen-3-tts-voice-design-1-7b/v1/qwen-3-tts-voice-design-1-7b-request"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "text": "The old clock tower chimed midnight as the fog rolled in from the harbor, wrapping the cobblestone streets in a blanket of grey.",
    "prompt": "Speak in a mysterious and atmospheric tone.",
    "language": "English"
}

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

const url = "https://gateway.pixazo.ai/qwen-3-tts-voice-design-1-7b/v1/qwen-3-tts-voice-design-1-7b-request";

const headers = {
  "Content-Type": "application/json",
  "Cache-Control": "no-cache",
  "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
};

const data = {
  text: "The old clock tower chimed midnight as the fog rolled in from the harbor, wrapping the cobblestone streets in a blanket of grey.",
  prompt: "Speak in a mysterious and atmospheric tone.",
  language: "English"
};

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/qwen-3-tts-voice-design-1-7b/v1/qwen-3-tts-voice-design-1-7b-request" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "text": "The old clock tower chimed midnight as the fog rolled in from the harbor, wrapping the cobblestone streets in a blanket of grey.",
    "prompt": "Speak in a mysterious and atmospheric tone.",
    "language": "English"
  }'

Output

{
  "request_id": "qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-3-tts-voice-design-1-7b_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.

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": "qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-3-tts-voice-design-1-7b",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.wav"
    ],
    "media_type": "audio/wav"
  },
  "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": "qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-3-tts-voice-design-1-7b",
  "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 - Qwen 3 TTS Voice Design [1.7B] generate request

Field	Type	Required	Default	Description
text	string	Yes	—	The spoken content to be synthesized into speech.
prompt	string	Yes	—	A style or tone description guiding the voice expression (e.g., "speak in a mysterious tone").
language	string	Yes	—	The language of the input text (e.g., "English", "Spanish").
top_k	integer	No	50	Controls diversity by limiting sampling to the top K most likely tokens.
top_p	float	No	1.0	Nucleus sampling threshold; lower values focus on higher-probability tokens.
temperature	float	No	0.9	Controls randomness; higher values increase creativity, lower values increase determinism.
repetition_penalty	float	No	1.05	Penalizes repeated tokens to reduce redundancy in output.
subtalker_dosample	boolean	No	true	Enables sampling for subtalker generation; disables greedy decoding.
subtalker_top_k	integer	No	50	Top-K sampling parameter for subtalker voice modeling.
subtalker_top_p	float	No	1.0	Nucleus sampling threshold for subtalker voice modeling.
subtalker_temperature	float	No	0.9	Temperature for subtalker voice modeling.
max_new_tokens	integer	No	200	Maximum number of tokens to generate in the output sequence.

Minimum Request

{
  "text": "The old clock tower chimed midnight as the fog rolled in from the harbor, wrapping the cobblestone streets in a blanket of grey.",
  "prompt": "Speak in a mysterious and atmospheric tone.",
  "language": "English"
}

Full Request (all options)

{
  "text": "The old clock tower chimed midnight as the fog rolled in from the harbor, wrapping the cobblestone streets in a blanket of grey.",
  "prompt": "Speak in a mysterious and atmospheric tone.",
  "language": "English",
  "top_k": 50,
  "top_p": 1.0,
  "temperature": 0.9,
  "repetition_penalty": 1.05,
  "subtalker_dosample": true,
  "subtalker_top_k": 50,
  "subtalker_top_p": 1.0,
  "subtalker_temperature": 0.9,
  "max_new_tokens": 200
}

Response

{
  "request_id": "qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	Your API subscription key

Response Handling

Common status codes for Qwen 3 TTS Voice Design [1.7B] generate request.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
403	Forbidden
404	Not Found
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 'qwen-3-tts-voice-design-1-7b' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-3-tts-voice-design-1-7b",
  "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/qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-3-tts-voice-design-1-7b",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen-3-tts-voice-design-1-7b_019dxxxx-xxxx/output.ext"
    ],
    "media_type": "application/octet-stream"
  },
  "created_at": "2026-03-31T10:00:00.000Z",
  "updated_at": "2026-03-31T10:00:15.000Z",
  "completed_at": "2026-03-31T10:00:15.000Z"
}

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.

Qwen 3 TTS Voice Design API Pricing

Your request will cost $0.09 per 1,000 characters.

describe a voice in words and generate speech in it

equivalent to $90 per 1M characters

9. Qwen 3 TTS Clone Voice 0.6B

Qwen 3 TTS Clone Voice 0.6B API Documentation

https://gateway.pixazo.ai/qwen-3-tts-clone-voice-0-6b/v1/qwen-3-tts-clone-voice-0-6b-request

Authentication

All requests require an API key passed via header.

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

Qwen 3 TTS Clone Voice [0.6B] generate request - Qwen 3 TTS Clone Voice [0.6B]

Request Code

POST https://gateway.pixazo.ai/qwen-3-tts-clone-voice-0-6b/v1/qwen-3-tts-clone-voice-0-6b-request
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "reference_text": "Welcome to the presentation. Today we will explore the fascinating world of artificial intelligence and its impact on everyday life."
}

import requests

url = "https://gateway.pixazo.ai/qwen-3-tts-clone-voice-0-6b/v1/qwen-3-tts-clone-voice-0-6b-request"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
    "reference_text": "Welcome to the presentation. Today we will explore the fascinating world of artificial intelligence and its impact on everyday life."
}

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

const url = "https://gateway.pixazo.ai/qwen-3-tts-clone-voice-0-6b/v1/qwen-3-tts-clone-voice-0-6b-request";
const headers = {
  "Content-Type": "application/json",
  "Cache-Control": "no-cache",
  "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
};
const data = {
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "reference_text": "Welcome to the presentation. Today we will explore the fascinating world of artificial intelligence and its impact on everyday life."
};

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/qwen-3-tts-clone-voice-0-6b/v1/qwen-3-tts-clone-voice-0-6b-request" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
    "reference_text": "Welcome to the presentation. Today we will explore the fascinating world of artificial intelligence and its impact on everyday life."
  }'

Output

{
  "request_id": "qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-3-tts-clone-voice-0-6b_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.

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": "qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-3-tts-clone-voice-0-6b",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.wav"
    ],
    "media_type": "audio/wav"
  },
  "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": "qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-3-tts-clone-voice-0-6b",
  "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 - Qwen 3 TTS Clone Voice [0.6B] generate request

Field	Type	Required	Default	Description
audio_url	string	Yes	—	An audio file containing the voice sample to clone. Supported formats: MP3, WAV, FLAC.
reference_text	string	Yes	—	The exact spoken text from the audio sample. This is used to align the audio with the transcript for accurate voice embedding.

Minimum Request

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "reference_text": "Welcome to the presentation. Today we will explore the fascinating world of artificial intelligence and its impact on everyday life."
}

Full Request (all options)

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "reference_text": "Welcome to the presentation. Today we will explore the fascinating world of artificial intelligence and its impact on everyday life."
}

Response

{
  "request_id": "qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	Your API subscription key

Response Handling

Common status codes for Qwen 3 TTS Clone Voice [0.6B] generate request.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
403	Forbidden
404	Not Found
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 'qwen-3-tts-clone-voice-0-6b' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-3-tts-clone-voice-0-6b",
  "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/qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-3-tts-clone-voice-0-6b",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen-3-tts-clone-voice-0-6b_019dxxxx-xxxx/output.ext"
    ],
    "media_type": "application/octet-stream"
  },
  "created_at": "2026-03-31T10:00:00.000Z",
  "updated_at": "2026-03-31T10:00:15.000Z",
  "completed_at": "2026-03-31T10:00:15.000Z"
}

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.

Qwen 3 TTS Clone Voice 0.6B API Pricing

Your request will cost $0.0007 per minute of reference audio.

$0.0007 per minute of the reference clip you upload

10. Qwen 3 TTS Clone Voice 1.7B

Qwen 3 TTS Clone Voice 1.7B API Documentation

https://gateway.pixazo.ai/qwen-3-tts-clone-voice-1-7b/v1/qwen-3-tts-clone-voice-1-7b-request

Authentication

All requests require an API key passed via header.

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

Qwen 3 TTS Clone Voice [1.7B] generate request - Qwen 3 TTS Clone Voice [1.7B]

Request Code

POST https://gateway.pixazo.ai/qwen-3-tts-clone-voice-1-7b/v1/qwen-3-tts-clone-voice-1-7b-request
Content-Type: application/json
Cache-Control: no-cache
Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "reference_text": "The quarterly results exceeded expectations across all major business segments, demonstrating strong momentum heading into the new fiscal year."
}

import requests

url = "https://gateway.pixazo.ai/qwen-3-tts-clone-voice-1-7b/v1/qwen-3-tts-clone-voice-1-7b-request"
headers = {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
}
data = {
    "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
    "reference_text": "The quarterly results exceeded expectations across all major business segments, demonstrating strong momentum heading into the new fiscal year."
}

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

const url = "https://gateway.pixazo.ai/qwen-3-tts-clone-voice-1-7b/v1/qwen-3-tts-clone-voice-1-7b-request";

const data = {
  audio_url: "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  reference_text: "The quarterly results exceeded expectations across all major business segments, demonstrating strong momentum heading into the new fiscal year."
};

fetch(url, {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "Cache-Control": "no-cache",
    "Ocp-Apim-Subscription-Key": "YOUR_SUBSCRIPTION_KEY"
  },
  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/qwen-3-tts-clone-voice-1-7b/v1/qwen-3-tts-clone-voice-1-7b-request" \
  -H "Content-Type: application/json" \
  -H "Cache-Control: no-cache" \
  -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" \
  --data-raw '{
    "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
    "reference_text": "The quarterly results exceeded expectations across all major business segments, demonstrating strong momentum heading into the new fiscal year."
  }'

Output

{
  "request_id": "qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-3-tts-clone-voice-1-7b_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.

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": "qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-3-tts-clone-voice-1-7b",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/output.wav"
    ],
    "media_type": "audio/wav"
  },
  "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": "qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-3-tts-clone-voice-1-7b",
  "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 - Qwen 3 TTS Clone Voice [1.7B] generate request

Field	Type	Required	Default	Description
audio_url	string	Yes	—	An audio file containing the speaker's voice. Supported formats: MP3, WAV, FLAC.
reference_text	string	Yes	—	The exact spoken text from the audio sample. Must match the content of the audio clip to enable accurate voice embedding.

Minimum Request

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "reference_text": "The quarterly results exceeded expectations across all major business segments, demonstrating strong momentum heading into the new fiscal year."
}

Full Request (all options)

{
  "audio_url": "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/doc-assets/audio/speech-17s.mp3",
  "reference_text": "The quarterly results exceeded expectations across all major business segments, demonstrating strong momentum heading into the new fiscal year."
}

Response

{
  "request_id": "qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "QUEUED",
  "polling_url": "https://gateway.pixazo.ai/v2/requests/status/qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Request Headers

Header	Value
Content-Type	application/json
Cache-Control	no-cache
Ocp-Apim-Subscription-Key	Your API subscription key

Response Handling

Common status codes for Qwen 3 TTS Clone Voice [1.7B] generate request.

Code	Meaning
202	Accepted — Request queued
400	Bad Request
401	Unauthorized
403	Forbidden
404	Not Found
429	Too Many Requests
500	Internal Server Error

Request Parameters - Qwen 3 TTS Clone Voice [1.7B] check status

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 'qwen-3-tts-clone-voice-1-7b' not found or is disabled"
}

Error via Status/Webhook

{
  "request_id": "qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "ERROR",
  "model_id": "qwen-3-tts-clone-voice-1-7b",
  "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/qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"

Response (Completed)

{
  "request_id": "qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
  "status": "COMPLETED",
  "model_id": "qwen-3-tts-clone-voice-1-7b",
  "error": null,
  "output": {
    "media_url": [
      "https://pub-582b7213209642b9b995c96c95a30381.r2.dev/v1/qwen-3-tts-clone-voice-1-7b_019dxxxx-xxxx/output.ext"
    ],
    "media_type": "application/octet-stream"
  },
  "created_at": "2026-03-31T10:00:00.000Z",
  "updated_at": "2026-03-31T10:00:15.000Z",
  "completed_at": "2026-03-31T10:00:15.000Z"
}

Response Fields

Field	Type	Required	Default	Description
request_id	string	Yes	—	Unique identifier returned from the initial request submission.

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.

Qwen 3 TTS Clone Voice 1.7B API Pricing

Your request will cost $0.0008 per minute of reference audio.

$0.0008 per minute of the reference clip you upload

⚡ Performance

Live usage measured on Pixazo's gateway, split by model version. Generation time is how long a generation takes end-to-end (lower is better). Success rate is the percent of generations that complete (higher is better).

Show data for the last

Generations

7,800last 30d

~260 per day

Success rate

76.9%

of completed generations

Generation time

31.9savg

p95 36.6s

Requests

Jul 26max 1,300Aug 24

Qwen 3.0 ASR Flash RealtimeAvg 57/day

Qwen 3.0 ASR FlashAvg 43/day

Qwen Audio 3.0 TTS PlusAvg 33/day

Qwen 3.0 TTSAvg 27/day

OtherAvg 100/day

Generation Time

Jul 26max 58.8sAug 24

Qwen 3.0 ASR Flash RealtimeAvg 34.9s

Qwen 3.0 ASR FlashAvg 37.5s

Qwen Audio 3.0 TTS PlusAvg 25.9s

Qwen 3.0 TTSAvg 37.6s

OtherAvg 28.5s

Error Rate

Jul 26max 100.0%Aug 24

Qwen 3.0 ASR Flash RealtimeAvg 23.5%

Qwen 3.0 ASR FlashAvg 7.7%

Qwen Audio 3.0 TTS PlusAvg 0.0%

Qwen 3.0 TTSAvg 37.5%

OtherAvg 33.3%

〰 Uptime

Percent of generations that succeeded over the selected period, per model version.

Avg. Success Rate (30d)

76.92%

across all generations of this model family

Uptime

Jul 26max 100%Aug 24

Qwen 3.0 ASR Flash RealtimeAvg 76.47%

Qwen 3.0 ASR FlashAvg 92.31%

Qwen Audio 3.0 TTS PlusAvg 100.00%

Qwen 3.0 TTSAvg 62.50%

OtherAvg 66.67%