voicebox_generate
Generate speech from text using a voice profile. Returns an ID to poll for generation status.
Instructions
Generate speech from text using a voice profile.
Returns a GenerationResponse with status="generating" and an id to poll at voicebox_get_generation_status.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| seed | No | Reproducible seed (null = random). | |
| text | Yes | Text to speak (max 50000 chars). | |
| engine | No | TTS engine — qwen, qwen_custom_voice, luxtts, chatterbox, chatterbox_turbo, tada, kokoro. | qwen |
| instruct | No | Optional instruction prompt (max 500 chars). | |
| language | No | Language code — en, zh, ja, ko, de, fr, ru, pt, es, it, he, ar, da, el, fi, hi, ms, nl, no, pl, sv, sw, tr. | en |
| normalize | No | Normalize output audio volume. | |
| model_size | No | Model size — 1.7B, 0.6B, 1B, 3B. | 1.7B |
| profile_id | Yes | Voice profile ID (use voicebox_list_profiles to find). | |
| personality | No | When true, rewrites text in-character before TTS. | |
| crossfade_ms | No | Crossfade between chunks in ms (0-500). | |
| max_chunk_chars | No | Max chars per chunk for long text (100-5000). |
Output Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||