Providers

Browserbase

Markdown from /v1/fetch without a session. No screenshot over REST and a CDP URL for everything else. Needs BROWSERBASE_API_KEY
IDbrowserbase02 / 8api.browserbase.com · 3 of 13
Provider / browserbase

Browserbase

Markdown from /v1/fetch without a session, CDP once you open one.

Supports
3 of 13 operations
Scrape
one call
Needs
BROWSERBASE_API_KEY
Endpoint
api.browserbase.com
operations 3 / 13

Operations

  • scrapeone call
  • screenshotno
  • elementno
  • pdfno
  • crawlno
  • linksno
  • a11y treeno
  • extractno
  • searchno
  • sessionsone call
  • cdpone call
  • navigateno
  • evaluateno

Recorded

Came back
markdown
Size
209 characters
Time
1,028 ms

This domain is for use in documentation examples without needing permission. This is not a service, avoid relying on it for testing and monitoring purposes.

Access

Createcreate("browserbase")
CLIbrowsers scrape https://example.com -p browserbase
Toolprovider: "browserbase"

Browserbase's /v1/fetch renders a page and hands back markdown without opening a browser session, so scrape() is one call. In the recording it was the tidiest answer of the lot: two sentences and a proper markdown link.

That's about it over REST. No screenshot, no PDF, no crawl. screenshot() throws UnsupportedOperationError and points you at CDP. Browserbase is built around sessions you drive yourself, and createSession() gives you exactly that: an ID and a CDP URL.