Translate Audio

POST

audio

translations

curl https://gateway.bud.studio/v1/audio/translations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "file=@french_audio.mp3" \
  -F "model=my-translation-deployment" \
  -F "response_format=json"

from openai import OpenAI

client = OpenAI(base_url="https://gateway.bud.studio/v1", api_key="YOUR_API_KEY")

with open("french_audio.mp3", "rb") as f:
    translation = client.audio.translations.create(
        model="my-translation-deployment",
        file=f,
    )
print(translation.text)

{
  "text": "Hello, this is the translated English text."
}

curl https://gateway.bud.studio/v1/audio/translations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "file=@french_audio.mp3" \
  -F "model=my-translation-deployment" \
  -F "response_format=json"

from openai import OpenAI

client = OpenAI(base_url="https://gateway.bud.studio/v1", api_key="YOUR_API_KEY")

with open("french_audio.mp3", "rb") as f:
    translation = client.audio.translations.create(
        model="my-translation-deployment",
        file=f,
    )
print(translation.text)

{
  "text": "Hello, this is the translated English text."
}

The request and response follow OpenAI’s audio/translations API, so OpenAI SDKs work unchanged. The output is always English. The model is the name of a speech-to-text deployment that supports translation.

Headers

Parameter	Type	Required	Description
Authorization	string	Yes	Bearer authentication header
Content-Type	string	Yes	`multipart/form-data`

Form Data

Parameter	Type	Required	Description
file	file	Yes	The audio to translate. The same files are accepted as for Transcribe Audio: up to 25 MB and 30 minutes, and no Opus.
model	string	Yes	Name of a deployment that supports translation. Any other deployment returns `404`; see Which deployments can translate.
response_format	string	No	`json`, `text`, `verbose_json`, `srt` or `vtt`. Default: `json`. The formats, their shapes and the subtitle segments are the same as for Transcribe Audio.
prompt	string	No	Text to guide the translation. Used by OpenAI, Groq and self-hosted deployments; any other provider ignores it with an `x-bud-config-warning`.
temperature	number	No	`0.0` to `1.0`; a value outside that range, or one that is not a number, is refused with `400`. Used by OpenAI, Groq and self-hosted deployments; any other provider ignores it with an `x-bud-config-warning`.
timestamp_granularities[]	array	No	`word`, `segment`, or both, for `verbose_json`. Same behaviour as on Transcribe Audio.

There is no language field: the source language is detected, and the target is always English.

Which deployments can translate

Translation is a capability of the deployment, not of the endpoint. A deployment of a provider that cannot translate returns 404 with the message Model '…' not found or does not support audio_translation, even though the same deployment works on Transcribe Audio. Providers that can translate an uploaded file to English are OpenAI (whisper-1) and Groq (with a model that supports translation, such as whisper-large-v3). Other providers, including Deepgram, ElevenLabs, AssemblyAI, Gladia and Speechmatics, do not translate uploads; use Transcribe Audio for them.

Response headers

Header	Description
Content-Type	Depends on `response_format`.
x-bud-config-warning	A request or deployment setting that could not be applied. Repeated once per warning; the request still succeeds.
x-bud-config-warning-count	How many `x-bud-config-warning` headers were sent.

Errors

Errors use OpenAI’s error format, and the status codes are the same as for Transcribe Audio.

404

{
  "error": {
    "message": "Model 'my-stt-deployment' not found or does not support audio_translation",
    "type": "invalid_request_error",
    "param": "model",
    "code": null
  }
}

curl https://gateway.bud.studio/v1/audio/translations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "file=@french_audio.mp3" \
  -F "model=my-translation-deployment" \
  -F "response_format=json"

from openai import OpenAI

client = OpenAI(base_url="https://gateway.bud.studio/v1", api_key="YOUR_API_KEY")

with open("french_audio.mp3", "rb") as f:
    translation = client.audio.translations.create(
        model="my-translation-deployment",
        file=f,
    )
print(translation.text)

{
  "text": "Hello, this is the translated English text."
}

Introduction

Gateway APIs

Python SDK & CLI

Headers

Form Data

Which deployments can translate

Response headers

Errors

​Headers

​Form Data

​Which deployments can translate

​Response headers

​Errors

Headers

Form Data

Which deployments can translate

Response headers

Errors