fastCRW is a search + scrape API for AI agents. It lets you search the web, scrape pages, crawl sites, map URLs, and return clean markdown, JSON, links, and metadata in one workflow. It’s open source, Firecrawl-compatible, benchmarked 2.3x faster than Tavily for search and 2.2x faster than Firecrawl for scraping, and runs as a lightweight single binary with 500 free credits to start.
Hey Product Hunt 👋
I built fastCRW because I kept seeing the same pattern:
AI products look great in demos, then fall apart in production when they need fresh web data.
Teams end up stitching together search, scraping, rendering, crawling, and extraction just to give an agent usable context. fastCRW is my attempt to make that simpler: search + scrape + crawl + map in one API, with clean markdown/JSON output, source URLs, MCP support, and an open-source core.
A big goal was making it easier to ship AI products people can actually trust — not just something that works on the happy path.
A few things that make fastCRW different:
search + scrape in one workflow
Firecrawl-compatible API
open-source core + managed cloud
public benchmarks
500 free credits to try it
Would really love feedback on:
whether the positioning is clear
whether the docs / first-run experience feel smooth
what still feels confusing, weak, or missing
Thanks for checking it out 🙏
