YouTube captions feed search indexes, summarization tools, subtitle generators, and accessibility work. The annoying part is fetching them reliably when page markup changes every few months.
A transcript API returns caption data over HTTP so you are not maintaining scrapers. This guide covers what to look for, how FreeTranscriptAPI is wired, and how to call it from your stack. See the API reference for endpoint details.
What is a YouTube transcript API?
You send a video URL or ID. You get spoken text back, usually as JSON with per-line timestamps, language selection, and documented rate limits.
Compared to Puppeteer or HTML parsing, the API owns caption fetching, language fallback, and format normalization.
Core endpoints
FreeTranscriptAPI exposes three endpoints at api.freetranscriptapi.com/v1:
- GET /transcript returns the full transcript. Pass
video_urlor a video ID. Optional:langfor language andformatfor json or plain text. - GET /languages lists caption languages before you fetch.
- GET /health for uptime checks.
Quick start
No account required. Unauthenticated calls get 20 requests per hour per IP.
Example request
curl "https://api.freetranscriptapi.com/v1/transcript?video_url=https://www.youtube.com/watch?v=dQw4w9WgXcQ"Response format
Default JSON includes language, video title, and transcript segments. Each segment has text, start (seconds), and duration.
That shape drops straight into LLM prompts, subtitle tools, and search indexes. Pass format=text if you only need a string.
Rate limits and pricing
- No API key: 20 requests/hour per IP
- Free account: 1,000 credits on signup, no credit card
- Pay as you go: $1.29 per 1,000 credits (1 credit = 1 transcript lookup)
API vs scraping
Scraping with headless browsers or unofficial libraries works for one-off scripts. In production you inherit DOM churn, IP blocks, and captchas. Read why scraping breaks in production for the full picture.
Use an API when you need stable endpoints, published limits, JSON output, and language fallback without writing it yourself.
Common use cases
- Feeding transcripts into GPT or Claude for summarization pipelines
- Building searchable video libraries
- Turning podcast episodes on YouTube into blog drafts
- n8n, Zapier, or custom automation
- Subtitle or translation workflows
Compare free options in our free API roundup.