Skip to main content

Scraping using Browsr

Browsr ships quick extraction endpoints plus agent-friendly overlays so you can design and monitor scrapes without bolting on extra tooling.

Grab content as Markdown or HTML​

Use /browser/content when you already know the selector you want. The response contains only the rendered fragment, which keeps token usage low.

SESSION=session-123
SELECTOR=main

# Markdown (best for LLMs)
curl "http://localhost:8082/browser/content?session_id=$SESSION&selector=$SELECTOR&mode=markdown"

# HTML (alias: /browser/html)
curl "http://localhost:8082/browser/content?session_id=$SESSION&selector=$SELECTOR&mode=html"
  • selector targets any CSS selector.
  • mode supports markdown or html.
  • Use the same session_id you run commands against so cookies and navigation history stay intact.