TranscribeAI APIbeta

Quick start

The TranscribeAI API is included in the Unlimited plan. Create a key from the API button in the site header (when signed in), then send it in the Authorization header.

curl https://transcribeai.site/api/me \
  -H "Authorization: Bearer tai_YOUR_KEY"

Authentication and limits

• Header: Authorization: Bearer tai_YOUR_KEY. Keys start with tai_; keep them secret and revoke them from the site if exposed.
• Rate limit: 60 requests per minute per key. Over the limit you get 429 with a Retry-After header.
• Usage counts against your plan minutes (Unlimited: fair use of 80 hours per month). Files up to 60 minutes; uploads up to 1 GB.
• Up to 5 active keys per account. A key stops working if the account is no longer on Unlimited.

Transcribe a video link

POST /api/v1/transcribe. Public TikTok, Instagram, Facebook and X links are supported. YouTube links are not supported.

curl -X POST https://transcribeai.site/api/v1/transcribe \
  -H "Authorization: Bearer tai_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://www.tiktok.com/@user/video/123", "language": "fr"}'

Response (shortened):

{
  "transcript": "Bonjour à tous...",
  "detected_language": "english",
  "output_language": "fr",
  "translated": true,
  "platform": "tiktok",
  "duration_seconds": 42.5,
  "segments": [{"start": 0.0, "end": 3.2, "text": "Bonjour à tous"}],
  "minutes_used_this_period": 1.2,
  "minutes_limit": 4800
}

language is the output language. auto keeps the original spoken language. A code translates the transcript and its segments: en, fr, es, ar, de, it, pt, nl, ru, zh, ja, ko, tr, hi, pl. The response includes detected_language (spoken), output_language, translated (true/false) and, if something went wrong, translation_error.

Transcribe an uploaded file

POST /api/v1/transcribe-file. Send the file as the raw request body. Accepted: MP3, WAV, M4A, AAC, OGG, OPUS, FLAC, WMA, MP4, MOV, MKV, WEBM, AVI, M4V, 3GP. The language query parameter works as above (output language, auto = no translation).

curl -X POST "https://transcribeai.site/api/v1/transcribe-file?language=auto&filename=meeting.mp3" \
  -H "Authorization: Bearer tai_YOUR_KEY" \
  -H "Content-Type: application/octet-stream" \
  --data-binary @meeting.mp3

AI summary

POST /api/v1/summary

curl -X POST https://transcribeai.site/api/v1/summary \
  -H "Authorization: Bearer tai_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"transcript": "Full transcript text...", "language": "auto"}'

Response: {"summary": "..."}

Export subtitles and documents

POST /api/v1/export/{format} where format is txt, docx, pdf, srt or vtt. Pass the transcript and the segments returned by the transcription call.

curl -X POST https://transcribeai.site/api/v1/export/srt \
  -H "Authorization: Bearer tai_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"transcript": "Hello everyone", "segments": [{"start": 0.0, "end": 3.2, "text": "Hello everyone"}]}' \
  -o subtitles.srt

Use it in n8n or Python

n8n: add an HTTP Request node, method POST, URL https://transcribeai.site/api/v1/transcribe, header Authorization = Bearer tai_YOUR_KEY, body type JSON.

import requests

r = requests.post(
    "https://transcribeai.site/api/v1/transcribe",
    headers={"Authorization": "Bearer tai_YOUR_KEY"},
    json={"url": "https://www.tiktok.com/@user/video/123", "language": "auto"},
    timeout=300,
)
print(r.json()["transcript"])

Errors

Errors are JSON: {"detail": "message"}