MOSS-VoiceGenerator
LatestSpeech generationThe MOSS-VoiceGenerator API generates a speech preview from text to synthesize and voice description instructions. Describe the desired voice characteristics in natural language to generate speech in the corresponding style.
ENDPOINT
POST /v1/audio/voice/generations
Current request path
Input
Text + instruction
Content accepted by the request
Output
Audio
Content returned on success
Model overview
Generate speech from text and voice instructions. Use the stable model ID moss-voice-generator-1.0 by default; to reproduce a fixed version, pass moss-voice-generator-1.0-2026-07-22 in model.
Billing
Billing methodCNY priceActual deduction
By input characters
CNY 2.00 / 10K characters40 credits / 10K charactersEach request is billed by actual usage. CNY prices are for reference; the final charge follows the credits bill. View full pricing
Modality support
TTextInput only
AAudioOutput only
IImageNot supported
VVideoNot supported
Supported endpoints
Single-speaker speech
/v1/audio/speechNot supportedMulti-speaker dialogue speech
/v1/audio/speech/speakersNot supportedAudio transcription
/v1/audio/transcriptionsNot supportedVoice design
/v1/audio/voice/generationsSupportedMultimodal understanding
/v1/responsesNot supportedRealtime video understanding
/v1/realtimeNot supportedImage generation
/v1/images/generationsNot supportedCapabilities
Performance—
Inference speed—
Streaming Not supported
Async Supported
- Instruction-driven voice generation: describe the voice style in natural language
- Select the synchronous output format with response_format
- Return audio binary directly or JSON containing the result URL
- Supports asynchronous tasks; retrieve results through the task query endpoint
Versions and snapshots
Model ID
moss-voice-generator-1.0Snapshot
moss-voice-generator-1.0-2026-07-22Latest