Back to Developer Overview/API Reference

API Reference

OpenAPI 3.1

Machine-verified endpoint contracts, authentication requirements, schema specifications, and live SDK examples generated from the canonical OpenAPI specification.

openapi.json Try in Playground

POST/v1/transcriptions

Operation: createTranscription

Create Transcription

Submit spoken audio for multilingual transcription, speaker diarization, and verifiable citation extraction. ### Execution Behavior: - **Synchronous (`200 OK`)**: Audio files $\le$ 120 seconds ($\le$ 25 MB) are processed in real time and return complete speaker segments and evidence directly in the HTTP response. - **Asynchronous (`202 Accepted`)**: Files $>$ 120 seconds (up to 4 hours / 500 MB) or requests specifying `async=true` return a job ID immediately for status polling or automated webhook callbacks.

Request Parameters

Parameter	In	Type	Required	Description
Idempotency-Key	header	string	optional	Unique UUID v4 to prevent duplicate transcription execution and double-billing on network retries.

Request Body Content Types

multipart/form-dataapplication/json

Supports multipart form data with a direct file upload or JSON containing a public audio_url.

Integration Example

Multi-language SDK and HTTP implementation snippets for this endpoint.

createTranscription.sh

curl -X POST https://api.andebet.ai/v1/transcriptions \
  -H "Authorization: Bearer $ANDEBET_API_KEY" \
  -F "[email protected]" \
  -F "language=auto" \
  -F "diarization=true" \
  -F "evidence_timestamps=true"

Response Contract

Transcription completed synchronously. Returns full speaker dialogue and evidence citations.

createTranscription-200.json

200

{
  "$ref": "#/components/schemas/TranscriptionCompletedResponse"
}