Alibaba Cloud Model Studio:Integrate the speech synthesis model

Call a Token Plan speech synthesis model from an Agent tool, convert text to audio, and download the audio locally.

Supported models

Model	Capability
qwen-audio-3.0-tts-plus	Speech synthesis

For more models, see models supported by Token Plan.

Prerequisites

• A Token Plan subscription and an API key that starts with sk-sp-. For details, see Integration overview.
• Python 3.9 or later.
• Your AI coding tool supports Slash Commands. For configuration paths of other tools, see Other tools.

Interface

POST https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1/services/audio/tts/SpeechSynthesizer

Use Authorization: Bearer $ANTHROPIC_AUTH_TOKEN in the request header. The response contains an audio URL.

Step 1: Create a Slash Command

Configure the Token Plan API key as the environment variable $ANTHROPIC_AUTH_TOKEN.

Create .claude/commands/text-to-speech.md in the project root and add the following content:

Call the Token Plan speech synthesis API to convert text to an audio file and download it locally.

Text to synthesize: $ARGUMENTS

## Steps

1. Extract text, voice (default: longanhuan_v3.6), format (default: mp3), and sample_rate (default: 22050). If a model is specified, use that model; otherwise use qwen-audio-3.0-tts-plus.
2. Use the Bash tool to run the following command:
```bash
RESPONSE=$(curl -s -X POST "https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1/services/audio/tts/SpeechSynthesizer" \
  -H "Authorization: Bearer $ANTHROPIC_AUTH_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen-audio-3.0-tts-plus",
    "input": {
      "text": "<text>",
      "voice": "<voice>",
      "format": "<format>",
      "sample_rate": <sample_rate>
    }
  }')

AUDIO_URL=$(printf '%s' "$RESPONSE" | python3 -c 'import json,sys; print(json.load(sys.stdin)["output"]["audio"]["url"])')
curl -s -o "speech_$(date +%Y%m%d_%H%M%S).<format>" "$AUDIO_URL"
```
3. Output the path of the downloaded audio file.

Step 2: Synthesize speech

In Claude Code, enter /text-to-speech Hello, welcome to Model Studio. To change the voice, specify it in the command, for example /text-to-speech use longanyang to say: Hello.

Request parameters

Parameter	Required	Description
model	Yes	Speech synthesis model ID.
input.text	Yes	Text to synthesize.
input.voice	Yes	Voice, for example longanhuan_v3.6.
input.format	No	Audio format. Supports mp3, pcm, wav, and opus. Default: mp3.
input.sample_rate	No	Sample rate. Default: 22050.

For more parameters, see CosyVoice speech synthesis HTTP API. For available voices, see CosyVoice voice list.

Response

• Audio URL: output.audio.url.
• URL expiration time: output.audio.expires_at.
• Billed characters: usage.characters.
• Audio URLs expire. Download the audio promptly.

Limits

• The voice must match the model version. For available voices, see the voice list.
• Billing is based on valid characters. Actual Credits consumption is shown in Usage in the console.

Other tools

For extension mechanisms and configuration paths of other tools, see Integration overview. Save the Slash Command above to the corresponding path.