Firecrawl
Scrapes any site into LLM-ready markdown for RAG and agents
Overview
Firecrawl converts websites into clean markdown or structured JSON for RAG and agent pipelines, handling JavaScript rendering, full-site crawls, and search in one API built for wiring web data into LLMs.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 21 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- 1,000 credits/mo
- 2 concurrent requests
- Community support
- 5,000 credits/mo
- 5 concurrent requests
- Basic support
- 100,000 credits/mo
- 50 concurrent requests
- Standard support
- 500,000 credits/mo
- 100 concurrent requests
- Priority support
- 1,000,000 credits/mo
- 150 concurrent requests
- Priority support
- Custom credit volume
- Zero data retention, SSO
- Dedicated support with SLA
Use cases
What you can produce with Firecrawl
- Scrape API — converts single pages to clean markdown/JSON
- Crawl API — full-site crawling with sitemap discovery
- Search API — web search returning scraped, structured results
- /extract endpoint — LLM-driven structured data extraction
- MCP server for direct agent/IDE integration
- Native LangChain, LlamaIndex, and Dify connectors
- Open-source, self-hostable core under AGPL-3.0
- Zero data retention and SSO on the Enterprise plan
ASEAN Perspective
Firecrawl in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
Firecrawl is the default choice for turning messy web pages into markdown an LLM can actually use — it handles JS-heavy sites, full-site crawls, and now agentic extraction, and the open-source core (over 130,000 GitHub stars) means you can inspect or self-host the scraping logic instead of trusting a black box. Backed by a $14.5M Series A led by Nexus Venture Partners in August 2025, with Y Combinator and Shopify's Tobi Lütke among backers, and used by teams at Apple and Canva.
The catch is cost. Hacker News and Reddit threads are consistent on this: casual users hit the $16 Hobby plan's 5,000-credit ceiling fast, and the token-billed /extract endpoint for structured extraction can run up a bill quickly at scale. Self-hosting is possible but several users report the free version gets steadily less capable as new features ship cloud-only first. Good default for RAG ingestion; model your credit burn before committing to a paid tier.
What people say
Firecrawl's open-source repository has crossed 130,000 GitHub stars, by its own account one of the top 100 repos on GitHub, and the company says its SDKs pull 2.5 million-plus weekly downloads across npm and PyPI. That scale shows up in reviews: a G2 reviewer called it the "Best Scraper We've Used," and Product Hunt commenters consistently praise how reliably it turns unstructured pages into clean, structured output without a manual scraping stack.
Pricing is the recurring sore point. On Hacker News and Reddit, the same complaints surface repeatedly: "it is damn expensive," one user needing the $99/mo tier just to cover their usage, another calling it "egregiously expensive" once volume climbs past the free and Hobby tiers. The credit system charges per page for Scrape, Crawl and Map, but Stealth Mode runs 5 credits a page and the JSON-extraction endpoint bills by token, so extraction-heavy workloads can outrun expectations fast.
A second complaint cluster centers on the self-hosted, open-source edition. Multiple GitHub and Reddit threads describe it degrading in relative usefulness over time as new capabilities — search, agentic extraction — ship cloud-only first, with one user summarizing that the company "tries to push all users to pay now and make self-host useless." That's a common open-core tension, but worth knowing going in if self-hosting was the draw.
On the plus side, integration is frequently cited as a strength: native connectors for LangChain, LlamaIndex, and MCP mean it drops into existing agent stacks without much glue code, and the company claims 96% web coverage including JavaScript-heavy sites with P95 latency around 3.4 seconds across millions of scrapes. Firecrawl raised an oversubscribed $14.5M Series A in August 2025 led by Nexus Venture Partners, with Y Combinator and Shopify CEO Tobi Lütke participating, bringing total funding to $16.2M.
Net: strong technical reputation and heavy usage among AI teams doing RAG ingestion, tempered by a credit-based pricing model that surprises people who don't model usage upfront, and a self-host tier that's increasingly a demo rather than a real alternative to the paid API.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including Firecrawl's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Firecrawl unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Firecrawl directly →
Spotted something out of date? Suggest an update →
More in Code & Dev Tools
