Web Scraping & Search API for AI Agents | fastCRW

# fastcrw.com > fastCRW is a local-first web scraper, search, and crawl API built for AI agents. Small single Rust static binary, Firecrawl-compatible REST API, built-in MCP server. Public benchmark: 63.74% truth-recall (522 of 819 labeled URLs) — highest of three tools tested — ~92% scrape success (of reachable URLs), 0 errors. Self-host for free under AGPL-3.0 or use the managed cloud at fastcrw.com — the same API either way, with no exit cost. ## Key Facts - Public benchmark: 63.74% truth-recall (522 of 819 labeled URLs) — highest of three tools tested — ~92% scrape success (of reachable URLs), 0 errors — full latency distribution and one-command repro: https://fastcrw.com/benchmarks - Local-first: small single Rust static binary, slim Docker image, no external services required - No exit cost: identical API whether self-hosted (free under AGPL-3.0) or on the managed cloud - Firecrawl-compatible on overlap surface (`/v1/scrape`, `/v1/crawl`, `/v1/map`, `/v1/search`); see compatibility matrix in docs for divergence on response field names and error envelopes - AGPL-3.0 core, commercial license available - Search API: real-time `/v1/search` at 1 credit/query — about $0.47 per 1,000 searches on an annual Growth/Scale plan (~$0.55 monthly), the cheapest real-time web search API per 1,000 requests at that volume; full sourced price comparison: https://fastcrw.com/alternatives/cheapest-search-api. Search averaged 880 ms over a 100-query benchmark; the same call can optionally scrape each result (1 credit per scraped page) ## API Endpoints - [Scrape](https://docs.fastcrw.com/#scrape): POST /v1/scrape — extract markdown, HTML, links, or structured JSON from a single URL - [Crawl](https://docs.fastcrw.com/#crawl): POST /v1/crawl — async BFS crawl with job polling, up to 1000 pages - [Map](https://docs.fastcrw.com/#map): POST /v1/map — discover all URLs on a site via sitemap + link traversal - [Search](https://docs.fastcrw.com/#search): POST /v1/sear