webclaw

POST /v1/scrape

Get a YouTube transcript in one call.

Submit a supported YouTube URL to retrieve transcript text and video metadata when captions are available.

Feed a YouTube link to your LLM or agent and get a clean, ready-to-read transcript instead of raw captions. Send a watch, shorts, or youtu.be URL to /v1/scrape and webclaw returns a structured youtube block with title, channel, and duration, plus a top-level transcript field, alongside the standard payload.

View docs

What you get

Everything in one call.

Full transcript

A clean, readable transcript of the video returned as a single text field, ready to feed straight into an LLM.

Structured video block

A typed youtube object with title, channel, duration, and other metadata so you don't parse markdown to find them.

One endpoint, any URL

Watch, shorts, and youtu.be links all route to the same /v1/scrape call with no extra params or per-format handling.

LLM-ready by default

When captions are available, the response gives you transcript text without the video player interface.

How it works

From URL to output in four steps.

01

Send the URL

POST any youtube.com/watch, /shorts, or youtu.be link to /v1/scrape with your API key.

02

Auto-detect the video

webclaw recognizes the YouTube URL and pulls the captions and metadata instead of running the generic page extractor.

03

Build the payload

The transcript is assembled into one text field and the video details are shaped into a structured youtube block.

04

Transcript returned

You get back the youtube object and a top-level transcript field next to the standard scrape response in one JSON reply.

API

One request, structured back.

Video

Channel: 3Blue1Brown · Duration: 19 min

This is a three. It's sloppily writtenand rendered at an extremely lowresolution of 28x 28…

Full transcript returned in the transcript field.

FAQ

Frequently asked questions.

Book a call

Send the video URL to POST /v1/scrape with your API key. webclaw detects the YouTube link, pulls the captions, and returns a top-level transcript field plus a structured youtube block in one JSON response. No separate transcript endpoint or video ID parsing needed.

Ship an agent that actually sees the web.

One credit pool, every endpoint. Cancel anytime, or self-host the open-source core for free.

API docs