Audio Transcription

POST /v1/audio/transcriptions

Transcribe an audio file to text. Uses multipart/form-data.

Request

curl https://ghostmind.optdmsa.com/v1/audio/transcriptions \
  -H "Authorization: Bearer sk-gm-..." \
  -F "file=@audio.mp3" \
  -F "language=en" \
  -F "response_format=json"

Parameters

Field	Type	Required	Description
file	file	Yes	Audio file (multipart upload)
language	string	No	Language hint
response_format	string	No	json (default) or text
conversation_id	string	No	Link to an existing conversation

Response (JSON)

{
  "text": "Hello, this is a transcription of the audio file.",
  "language": "en",
  "duration": 5.2,
  "model": "whisper-1",
  "job_id": "transcription-job-uuid"
}

Response (Text)

When response_format=text, returns plain text:

Hello, this is a transcription of the audio file.

Privacy

• Transcript text is not stored in the database
• Audio files are deleted immediately after transcription
• Only metrics are recorded (mime type, size, language, duration, status)

Errors

Status	Error Type	Cause
400	audio_validation_failed	Invalid audio file
404	conversation_not_found	Invalid conversation_id
429	transcription_rate_limit	Rate limited
503	transcription_disabled	Transcription is disabled
503	transcription_provider_unavailable	No provider available
504	transcription_timeout	Transcription timed out

Next Steps

• Transcription Guide — User guide
• Speech Generation — TTS API