Extract structured data
from any website
Send a URL and a JSON Schema. Get back clean, structured data. Powered by LLM agents that crawl, read, and extract.
Built for developers
A simple API that handles the complexity of web scraping and data extraction.
Schema-Driven Extraction
Your JSON Schema is both the output format and the extraction instructions. The description fields tell the LLM what to look for.
3-Level Deep Crawl
Automatically crawls the target site up to 3 levels deep, then extracts data across all fetched pages.
Smart Rejection
If the website doesn't match your schema criteria, the API returns early with a rejection — saving tokens and time.
SSE Streaming
Long-running extractions? Use the streaming endpoint with heartbeat events to survive reverse proxy timeouts.
How it works
Define your schema
Write a JSON Schema with description fields as extraction instructions. Specify what's relevant and what to ignore.
POST to /extract
Send the target URL and your schema. The agent crawls the site, reads the content, and extracts data.
Get structured data
Receive clean JSON matching your schema — validated, structured, and ready to use.
Pay for what you use
Prepaid credit, charged per extraction against the tokens it actually used. No subscription, no monthly minimum, no per-seat anything.
free credit when you sign in, so you can try it before paying
hard cap per request, enforced mid-run — one page can never surprise you
you pay for the work done, not for a tier you did not fill