Hyperbrowser

IDhyperbrowser05 / 8api.hyperbrowser.ai · 7 of 13

Provider / hyperbrowser

Hyperbrowser

One fetch route for pages and screenshots, jobs for crawl and extract, and search.

Supports: 7 of 13 operations
Scrape: one call
Needs: HYPERBROWSER_API_KEY
Endpoint: api.hyperbrowser.ai

operations 7 / 13

Operations

scrapeone call
screenshotone call
elementno
pdfno
crawlone call
linksno
a11y treeno
extractone call
searchone call
sessionsone call
cdpone call
navigateno
evaluateno

Recorded

Came back: markdown
Size: 167 characters
Time: 605 ms

# Example Domain

Access

Createcreate("hyperbrowser")

CLIbrowsers scrape https://example.com -p hyperbrowser

Toolprovider: "hyperbrowser"

Hyperbrowser does a lot through /api/web/fetch: markdown for scrape(), a stored image for screenshot(), both without a session. It answered fastest in the recording. It also answered with an older example.com than everyone else, the one with a heading and a different link. Somebody's cache was having a nice day.

Crawl and extract are jobs. The package starts one, polls it for up to two minutes, and if it's still running hands back the job ID instead of giving up. See Crawl and search.

A page that fails to load comes back from Hyperbrowser as HTTP 200 with status: "failed". The package turns that into an error, so an empty data never passes for an empty page.