Crawl → Execute
AI-generatedSummary
Perform an AI-powered crawl of a specified URL to extract data from the webpage according to optional user instructions and formatting preferences.
Inputs
- URL (required) — The URL to crawl.
- User Prompt — Optional instructions guiding what information to extract from the page.
- Output Format — Format of the extracted data, either 'markdown' (default) or 'json'.
- Schema (JSON) — OpenAPI JSON schema for structured extraction, required if Output Format is 'json'.
- Render JavaScript — Whether to render JavaScript on the page during crawling (default: false).
- Return Sources Limit — Maximum number of source references to return, between 1 and 100 (default: 25).
- Geo Location — Geographic location string to localize the crawling experience.
Output shape
A single JSON object containing the extracted data from the crawled page, formatted as specified (markdown string or JSON object).
The operation polls the AI crawler service until completion or timeout; errors such as invalid schema or crawl failure are returned as execution errors or, if Continue On Fail is enabled, as error messages within the output data.