Read one page (scrape)
One page by URL, for drop-in migration from hosted scrape APIs. The same path as the unbrowse_scrape MCP tool: plain HTTP when the server's HTML carries the content, else Unbrowse's cloud browser renders it. Charged as one verified call only on success. It never runs past deadlineMs (default 45 s, clamped 5–120 s). Every answer carries metadata.statusCode / metadata.status (the page's real HTTP status, also when rendered), timing (ms, httpMs, renderMs, queuedMs) and attempts (HTTP tries); every error carries the same in error.details, with stage (http or render), reason, httpStatus and renderStatus.
Errors: 400 invalid input; 402 out of calls; 429 blocked (a bot check or an empty page over HTTP and in the browser) with a Retry-After header when the page is remembered as walled — for 10 minutes the same host, country and path answer 429 at once; 504 scrape_timeout when the deadline passed.
Request
- Base URL: https://unbrowse.ai
- URL: https://unbrowse.ai/api/v1/scrape
- Auth: one of:
- OAuth 2 (authorization code, authorization URL https://unbrowse.ai/authorize, token URL https://unbrowse.ai/oauth/token, scopes: unbrowse)
- HTTP bearer
Request body
Response
The page
Changes
No recorded changes to this operation across all 1 revision of this API.