Three steps to your first transcript via the API: request a key, attach it to every call, and start transcribing.
Keys aren't self-serve yet. Email us and we'll have one in your inbox within a day.
Pass it as an X-API-Key header. Required on all API calls.
| Header | When | Notes |
|---|---|---|
X-API-Key: <key>required |
every request | Identifies your app and tier. Keep it out of client-side code — proxy through your backend. |
Authorization: Bearer <idToken>optional |
signed-in users | Firebase ID token. Ties the job to the user's account — shows in "Your files", inherits their plan limits. |
X-User-Email: you@example.comoptional |
server-to-server | Attaches the job to a Whipscribe account by email. Use when Firebase auth isn't available (backend pipelines, cron jobs). |
POST a URL or upload a file. Poll the job until done. Pull the transcript in any format.
Submit a public audio or video URL:
curl -X POST https://whipscribe.com/api/v1/transcribe/url \ -H "X-API-Key: $WHIPSCRIBE_KEY" \ -H "Content-Type: application/json" \ -d '{"url":"https://example.com/audio.mp3","language":"auto"}'
Poll until status is done:
# Use job_id from the submit response curl https://whipscribe.com/api/v1/jobs/$JOB_ID \ -H "X-API-Key: $WHIPSCRIBE_KEY"
Pull the transcript in the format you want:
# formats: txt · json · srt · vtt · docx curl https://whipscribe.com/api/v1/jobs/$JOB_ID/transcript?format=srt \ -H "X-API-Key: $WHIPSCRIBE_KEY"
File uploads, batch jobs, speaker labels, word timestamps, webhooks, idempotency keys, and error codes — all in the docs.
Can I try the API without a key?
Yes. Anonymous calls (no X-API-Key) are allowed on the free tier and subject to per-IP rate limits. You won't get a persistent job history or higher-than-free throughput without a key.
What's the base URL?
https://whipscribe.com/api/v1. CORS is open for GET and POST from browser origins — useful for client-side prototypes. For production, proxy through your backend so the key stays server-side.
How is usage billed?
Credit is metered in audio-hours. A 30-minute file spends 0.5h. Jobs that fail don't consume credit. Submits over your balance return HTTP 402 with an upgrade_url. See pricing for tiers.
What audio and video formats are supported?
MP3, MP4, M4A, WAV, OGG, FLAC, WEBM, MOV, and most other common formats — ffmpeg under the hood. YouTube URLs, podcast RSS enclosures, and public CDN links are accepted by the URL endpoint.
Is my data used for AI training?
No. Whipscribe runs Whisper on-prem. Your audio is never sent to a third-party ML provider and is not used to train any model.
Can I get a higher rate limit or on-prem deployment?
Email contact@neugence.ai with your volume requirements. We accommodate high-throughput and air-gapped setups.