Skip to Content

POST /api/transcribe/async

Request: multipart (audio field) or raw binary body.

Response (200):

{ "task_id": "stt:..." }
curl -sS -X POST "https://llm.dat.ai/api/transcribe/async" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "audio=@/path/to/audio.mp3"

With optional fanout:

curl -sS -X POST "https://llm.dat.ai/api/transcribe/async?fanout=3" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "audio=@/path/to/audio.mp3"

POST /api/transcribe/sync

Same input as async. Blocks until transcription completes (timeout ~10 minutes).

Response (200):

{
  "task_id": "stt:...",
  "success": true,
  "payload": "transcribed text ...",
  "execution_time_ms": 12345
}
curl -sS -X POST "https://llm.dat.ai/api/transcribe/sync" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/octet-stream" \
  --data-binary @/path/to/audio.wav

GET /api/transcribe/status

Query: task_id (required)

curl -sS "https://llm.dat.ai/api/transcribe/status?task_id=stt:YOUR_TASK_ID" \
  -H "Authorization: Bearer YOUR_API_KEY"

Response (200):

{
  "task_id": "stt:...",
  "status": "completed",
  "created_at": "2026-01-13T12:34:56Z",
  "audio_duration_seconds": 42,
  "success": true,
  "payload": "..."
}
TranscribingInference