Creative
Betacreative_tts_generate
Start a text-to-speech job with a model and voice available to the connected workspace.
Operation: Write. Requires read-and-write OAuth consent.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
body | object | Yes | Speech generation input. |
body.model | string | Yes | Model ID from creative_tts_models_list. |
body.transcript | string | Yes | Text to read aloud. |
body.voice | object | Yes | Voice selection. |
body.voice.id | string | Yes | Voice ID from creative_tts_voices_list for the model’s voice_family. |
body.output_format.container | string | No | Choose mp3 or wav if the model supports it. |
body.mode | string | No | Use async to receive a job immediately. |
idempotency_key | string | Yes | Use a unique key per action. Reuse it to retry identical input. |
approval_receipt | string | No | Receipt returned with the plan. Supply it after user confirmation. |
Use creative_tts_models_get to check the voice family, supported settings, and pricing.
The first call returns a plan. After confirmation, resubmit the same input with its approval_receipt. The job calls an external provider and can consume Creative credits. Pass data.job.id to creative_jobs_wait, then retrieve the output with creative_jobs_result.
Sample prompt
Read “The next train is arriving” using a voice supported by the speech model, then return the audio URL.