Audio API

Text-to-speech (TTS) and voice list queries

The Audio API provides text-to-speech (TTS) and voice listing capabilities.

Endpoints

Method	Path	Description
GET	`/v1/audio/voice/{model}`	List voices for a TTS model
POST	`/v1/audio/speech`	Text-to-speech (TTS)

Recommended call order: First call the voice list endpoint to get data[].code, then pass it as the voice field in the TTS request.

GET `/v1/audio/voice/{model}`

List available voices for a TTS model.

Auth: Bearer Token

Path parameter: model — TTS model name, e.g. tts-1, cosyvoice-v3-flash

Request: No request body

Request example (curl)

curl https://www.guid.ai/v1/audio/voice/cosyvoice-v3-flash \
  -H "Authorization: Bearer sk-xxx"

Response body

HTTP 200

{
  "success": true,
  "object": "list",
  "data": [
    {
      "name": "龙小淳",
      "code": "longxiaochun"
    },
    {
      "name": "Cherry",
      "code": "Cherry"
    }
  ]
}

Response fields

Field	Type	Description
`success`	boolean	Whether the call succeeded
`object`	string	Always `list`
`data`	array	Voice list
`data[].name`	string	Voice display name
`data[].code`	string	Voice code; pass as `voice` in TTS requests

When no voices are available:

{
  "success": true,
  "object": "list",
  "data": []
}

Failure response

HTTP 400

{
  "success": false,
  "message": "model is required"
}

Field	Type	Description
`success`	boolean	Always `false`
`message`	string	Failure reason

Notes:

Channel voices: name equals code; comes from channel configuration
Platform voice library: name is the display name; code is voice_key. If the model supports private user voices, they are included as well

POST `/v1/audio/speech`

Text-to-speech (TTS).

Auth: Bearer Token

Content-Type: application/json

Request body

{
  "model": "tts-1",
  "input": "Hello, welcome to Guid AI text-to-speech service.",
  "voice": "alloy",
  "response_format": "mp3",
  "speed": 1.0
}

Request fields

Field	Type	Required	Description
`model`	string	Yes	TTS model, e.g. `tts-1`, `cosyvoice-v3-flash`
`input`	string	Yes	Text to synthesize
`voice`	string	Yes	Voice code (see voice list above)
`response_format`	string	No	Output format: `mp3` (default), `opus`, `aac`, `flac`, `wav`, `pcm`
`speed`	number	No	Speech rate, 0.25–4.0; default 1.0

CosyVoice model example:

{
  "model": "cosyvoice-v3-flash",
  "input": "你好，欢迎使用语音合成服务。",
  "voice": "longxiaochun",
  "response_format": "mp3",
  "speed": 1.0
}

Request example (curl)

curl https://www.guid.ai/v1/audio/speech \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "tts-1",
    "input": "Hello, welcome to Guid AI text-to-speech service.",
    "voice": "alloy",
    "response_format": "mp3",
    "speed": 1.0
  }' \
  -o speech.mp3

CosyVoice model:

curl https://www.guid.ai/v1/audio/speech \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "cosyvoice-v3-flash",
    "input": "你好，欢迎使用语音合成服务。",
    "voice": "longxiaochun",
    "response_format": "mp3",
    "speed": 1.0
  }' \
  -o speech.mp3

Response body

HTTP 200

Content-Type: audio/mpeg (mp3) or the MIME type for the chosen format
Body: Audio binary stream (not JSON)

Example response headers:

HTTP/1.1 200 OK
Content-Type: audio/mpeg
Content-Length: 45632

Response notes

Item	Description
HTTP status	`200`
Content-Type	Determined by `response_format`; see table below
Body	Audio binary stream, not JSON

`response_format` and Content-Type mapping

`response_format`	Content-Type
`mp3` (default)	`audio/mpeg`
`opus`	`audio/opus`
`aac`	`audio/aac`
`flac`	`audio/flac`
`wav`	`audio/wav`
`pcm`	`audio/pcm`

Failure response

HTTP 400. Format: Standard errors.

{
  "error": {
    "message": "voice is required",
    "type": "invalid_request_error",
    "code": "invalid_request"
  }
}

Videos API

Create, poll, and download asynchronous video tasks

MCP Connection & Configuration

How to connect to and configure the Guid AI MCP service