No enabled audio-translation model is currently documented. This page intentionally does not present a runnable model ID.
@premai/api-sdk 1.0.59 and the OpenAI-compatible proxy contain an audio-translation method and /v1/audio/translations route. A route existing in the package is not sufficient to make a translation example operational: the account must also receive an enabled translation-capable model from the model catalogue.
The August 4 catalogue did not return openai/whisper-large-v3. Do not use it as a runnable example.
Check availability
List the live models and filter by type before implementing translation:
A runnable translation example requires an enabled audio-translation model, a documented response contract, and a successful authenticated test with a representative audio file. No model currently meets those conditions in this guide.
For the currently enabled transcription path, see Audio Transcription.
