generate_audio
Generate audio. TTS (default elevenlabs-tts): speak text in a named voice (jessica, sarah, liam, chris, roger, lily, bill, matilda, or a raw ElevenLabs voice id), 16 credits per 1,000 characters, rounded up. Emotion tags are honoured. TTS cloned voices cannot speak Hebrew: a stock voice of the same gender speaks that line. Gemini TTS (model: gemini-tts, Gemini 3.8 Flash TTS): the natural voice for Hebrew and any language; voice sulafat (female, default) or achird (male), optional language (ISO code, e.g. he) and direction (how to read it, e.g. "warm, natural, conversational, like telling a friend"; never read aloud; no emotion tags or tone), 5 credits per 1,000 characters, rounded up; returns an MP3 you can pass as audio_refs to generate_video. Cartesia TTS (model: cartesia-tts, Cartesia Sonic 3.6): native Hebrew voices for narration, 44 languages; voice ayala (female narrator, default), yael, shira, hila (female), gil, noam, yonatan (male), or a Cartesia voice id; Hebrew text speaks Hebrew (optional language, ISO code); optional speed (0.6 to 1.5) and emotion (calm, sad, excited, ...; best in English); no tone or direction, 13 credits per 1,000 characters, rounded up; returns an MP3 you can pass as audio_refs. Voice conversion (model: elevenlabs-sts): pass audio_url (public HTTPS audio, max 15 seconds) and voice to re-voice an existing performance into the locked voice. Optional stability, similarity_boost, style (0..1), seed (uint32), remove_background_noise (boolean). No text or tone for STS. STS costs 1 credits per measured input second, total rounded up; call quote first. Models: elevenlabs-tts, elevenlabs-sts, gemini-tts, cartesia-tts. Returns a job id; poll get_run_status for the audio URL (TTS MP3, STS WAV with source duration).
RETRY SAFETY: set request_id to a fresh UUID for a new generation. Reuse the exact same request_id and inputs only to recover a lost response; that returns the original job without another charge. A deliberately new generation needs a new request_id. Keep the returned job_id and poll get_run_status. Never invent a refund.
Tool call template
This tool declares no required parameters. Add optional parameters from the input schema as needed.
{
"id": 1,
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"arguments": {},
"name": "generate_audio"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| seed | No | STS only: best-effort deterministic seed. | |
| text | No | Required for TTS; forbidden for elevenlabs-sts. The words to speak. Emotion tags like [excited] or [whispers] are honoured by elevenlabs-tts (gemini-tts: use direction instead). | |
| tone | No | elevenlabs-tts only: energetic | calm | confident | dramatic. | |
| model | No | A live audio model id from list_models, or "auto". Omit for the default (elevenlabs-tts). gemini-tts: Gemini 3.8 Flash TTS, natural Hebrew and any language. cartesia-tts: Cartesia Sonic 3.6, native Hebrew voices for narration. | |
| speed | No | cartesia-tts only: speaking speed, 0.6 to 1.5 (1 is normal). | |
| style | No | STS only: style exaggeration, 0..1. | |
| voice | No | elevenlabs-tts / elevenlabs-sts: stock voice name or raw ElevenLabs voice id (required for STS; TTS defaults to sarah; TTS cloned voices cannot speak Hebrew: a stock voice of the same gender speaks a Hebrew line). gemini-tts: sulafat (female, default) or achird (male). cartesia-tts: ayala (female narrator, default), yael, shira, hila (female), gil, noam, yonatan (male), or a Cartesia voice id. | |
| emotion | No | cartesia-tts only: the emotion to read it with (neutral, calm, content, sad, angry, scared, excited, ...). Guidance; works best in English. | |
| language | No | gemini-tts and cartesia-tts only: the spoken language as an ISO code (he, es, pt-BR) for native pronunciation. cartesia-tts: Hebrew text defaults to he. | |
| audio_url | No | Required only for elevenlabs-sts: public audio URL, at most 15 seconds. Duration is measured by the server. | |
| direction | No | gemini-tts only: how to read the line, e.g. "warm, natural, conversational, like telling a friend". Never read aloud. | |
| stability | No | STS only: voice stability, 0..1. | |
| request_id | No | Unique request identity (prefer a UUID). Reuse with identical inputs to recover the same job after a lost response; choose a new identity only for an intentionally new generation. | |
| similarity_boost | No | STS only: voice similarity, 0..1. | |
| remove_background_noise | No | STS only: remove background noise from the input. |