Every API, one clean contract.

Fetch one URL → AI-ready content in the requested formats (markdown/text/html/rawHtml/metadata/links/images/). + trafilatura; honest extraction confidence + render flag. Firecrawl /scrape + Jina Reader parity.

{
  "url": "https://en.wikipedia.org/wiki/Web_scraping",
  "formats": [
    "markdown",
    "metadata"
  ]
}

Discover a site's URL surface: robots.txt + sitemap.xml(+index) + homepage internal links → deduped URL list with per-URL page-type hints (home/pricing/docs/blog/product/legal/contact/about). Firecrawl /map parity + page-type classification (our edge).

{
  "url": "https://stripe.com",
  "limit": 50
}

SYNC-bounded crawl from a seed URL (max 25 pages): BFS over same-domain internal links with depth + include/exclude regex, scraping each page. Large crawls are a phase-2 async job (not promised here). Firecrawl /crawl parity (bounded).

{
  "url": "https://example.com",
  "max_pages": 2,
  "max_depth": 1
}

Deterministic structured extraction from one URL: objects + microdata/RDFa presence + tables (parsed rows) + heading outline + prices + emails/phones, plus an optional {field: jsonld_dotpath} schema map. LLM-assisted extraction = phase-2 flag.

{
  "url": "https://en.wikipedia.org/wiki/Python_(programming_language)"
}

Scrape up to 10 URLs concurrently in one call (each SSRF-guarded). Shared formats. Firecrawl /batch-scrape parity (bounded).

{
  "urls": [
    "https://example.com",
    "https://en.wikipedia.org/wiki/HTML"
  ]
}