us/crw

Overview

A Rust-based web scraper, crawler, and search API that provides a drop-in compatible replacement for Firecrawl. It includes an MCP server for AI agents and is designed to be fast and lightweight. Benchmarks show it is 2.3x faster than Tavily and 1.5x faster than Firecrawl while using only 6 MB of RAM.

Best for

Best for
Developers needing a fast, lightweight self-hosted alternative to Firecrawl or Tavily for AI agent workflows

Use cases

• Building a custom search engine for an AI agent
• Scraping and crawling web pages at scale with low memory footprint
• Replacing Firecrawl or Tavily with a faster self-hosted alternative

How to use

Install

npx skills add us/crw

Tools exposed

• crw-mcp

Tested with

Claude Code, Cursor, Windsurf, Cline, Continue.dev, Codex, Gemini CLI

Example client config

export CRW_API_KEY="crw_live_..."\nexport CRW_API_URL="https://api.fastcrw.com"

Notes

A Rust-based web scraper, crawler, and search API that provides a drop-in compatible replacement for Firecrawl. It includes an MCP server for AI agents and is designed to be fast and lightweight. Benchmarks show it is 2.3x faster than Tavily and 1.5x faster than Firecrawl while using only 6 MB of RAM.

321 stars on GitHub. Last updated 2026-07-12. Licensed AGPL-3.0.

Use cases

• Building a custom search engine for an AI agent
• Scraping and crawling web pages at scale with low memory footprint
• Replacing Firecrawl or Tavily with a faster self-hosted alternative

Pros

• Very fast and memory efficient (6 MB RAM single binary)
• Drop-in compatible with Firecrawl API endpoints
• Supports self-hosting or managed cloud option

Cons

• Newer project with smaller community and fewer integrations
• May lack some advanced features of more established alternatives
• Rust binary may be less accessible for non-Rust developers

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.