Docs

Batch API

The batch API is the cheapest way to transcribe long files and archives. Submit a job, then poll it or wait for a webhook.

Distributed processing. Jobs run on a network of vetted consumer GPUs and Macs owned by independent contributors. Do not send personal or sensitive data. Read the data policy.

Create a job

curl https://api.vouchcompute.com/v1/audio/transcriptions/jobs \
  -H "Authorization: Bearer $VOUCH_API_KEY" \
  -H "Idempotency-Key: 0d4f1c3e-…" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "whisper-large-v3-turbo",
    "url": "https://example.org/archive/episode-42.mp3",
    "language": "en",
    "response_format": "srt",
    "webhook_url": "https://hooks.example.org/vouch",
    "metadata": {"episode": "42"}
  }'

Response 202:

{ "id": "trj_…", "object": "transcription.job", "status": "queued", "created_at": "2026-11-18T10:00:00Z", "estimated_units": "2730.000" }

URL rules: https only, public hosts only (private, loopback and link-local addresses are refused with input_url_rejected), at most 3 redirects, 2 GB and 4 hours of audio.

Track it

GET /v1/jobs/{id} returns status: queued, processing, completed, failed or cancelled. When completed, units is the billed audio seconds.

Get the result

GET /v1/jobs/{id}/result redirects to a signed download URL valid for 7 days, or returns the body inline when it's 1 MB or less.

Cancel

POST /v1/jobs/{id}/cancel is idempotent. Work not yet delivered is not charged.

Python

import httpx, uuid

job = httpx.post(
    "https://api.vouchcompute.com/v1/audio/transcriptions/jobs",
    headers={"Authorization": f"Bearer {key}", "Idempotency-Key": str(uuid.uuid4())},
    json={"model": "whisper-large-v3-turbo", "url": url, "response_format": "vtt"},
).json()
Batch API · Vouch Compute