Transcribe audio

Transcribe audio to text. Drop-in replacement for OpenAI /v1/audio/transcriptions.

Accepts multipart/form-data with an audio file. Supported models: whisper-1, google/gemini-2.5-flash.

Accepts tri-mode auth (user token / service+on-behalf-of / service-only) — the service-key path is required for backend callers like BlinkLife that proxy STT on a user's behalf, mirroring /v1/audio/speech.

POST

https://truememory-api.mvstg.com/v1/audio/transcriptions

View as Markdown

Authorization

HTTPBearerHTTPrequired

Header parameters

X-On-Behalf-Ofoptional
One of the following:
string
null
}
X-External-User-Idoptional
One of the following:
string
null
}
X-App-Idoptional
One of the following:
string
null
}

Request bodyFORM-DATArequired

filestringrequired
modeloptionalstring
default: "whisper-1"
languageoptional
One of the following:
string
null
}
promptoptional
One of the following:
string
null
}
response_formatoptionalstring
default: "json"
temperatureoptionalnumber
default: 0

Returns

200

The transcription. JSON by default; set response_format to text, srt or vtt for those formats instead.

JSON

One of the following:

object
string

422

Validation Error

JSON

detailoptionalarray ofobject
- locarray ofrequired
  One of the following:
  string
  integer
- msgstringrequired
- typestringrequired
- inputoptionalunknown
- ctxoptionalobject
}

curl --request POST \ --url https://truememory-api.mvstg.com/v1/audio/transcriptions \ --header 'Authorization: Bearer <token>' \ --form-string 'file=string' \ --form-string 'model=whisper-1' \ --form-string 'language=string' \ --form-string 'prompt=string' \ --form-string 'response_format=json' \ --form-string 'temperature=0'