The developer-friendly API for extracting reliable caption data from public YouTube videos. JSON, plain text, SRT, and VTT — one dependable endpoint.
< 1 sec
warm response time
100+
caption languages
JSON · SRT · VTT
formats ready to use
99.9%
uptime target
One API, no cleanup
Built for products that care about clean data and developers who don’t want to maintain another fragile scraper.
GET /v1/transcript?url=dQw4w9WgXcQ Authorization: Bearer ts_live_... { "language": "en", "segments": [...], "text": "Most weeks, I start by..." }
Warm requests resolve in under a second. Repeat lookups hit a cache and are effectively instant.
Timestamped segments, video metadata, and language info in one response — no HTML scraping on your end.
Get subtitle-ready files directly, formatted and gap-filled, for video and caption pipelines.
Works with human-written subtitles and YouTube's auto-generated captions, across 100+ languages.
Pass a language code to fetch a specific translation, or let it default to the original track.
We run our own extraction pipeline rather than reselling someone else's flaky wrapper.
Pricing
Start free. Upgrade when you need more volume — no credit card required to try the API.
One-time trial allowance.
For small scripts and side projects.
For side projects and small tools.
For production apps.
For scale and teams.
Frequently asked questions
Use JSON for structured data, plain text for reading and analysis, or SRT and VTT for subtitle workflows.
Yes. Request a BCP-47 language code when available, or use the default track for a straightforward integration.
No. Public videos may have no caption tracks, only auto-generated captions, or no track in the requested language.
No. TranscriptYT is an independent service for working with public caption tracks and is not affiliated with YouTube.