About Firecrawl
Overview
Firecrawl is a web data API that turns the live internet into clean Markdown and structured data for large language models, AI agents, and AI-native software. Developers call its endpoints to search the web with full page content, scrape any URL, crawl and map sites, and interact with pages inside cloud browser sandboxes.
The company reports more than 150,000 companies and 1.25 million developers using the platform, and its open-source repository has passed 180,000 GitHub stars, placing it among the top 100 repositories on GitHub. Firecrawl started in 2022 through Y Combinator's S22 batch, and the founding team previously built Mendable and SideGuide.
Key Benefits
- One API covers search, scrape, crawl, map, and browser interaction, so developers skip stitching together proxies, scrapers, and post-processing logic.
- Scraped pages come back as clean Markdown, structured JSON, screenshots, or semantic text, with navigation, footers, and ads stripped out.
- Firecrawl reports 96% coverage on its own 1,000-URL scrape benchmark and a P95 latency of 3,387 ms for real-time agent workloads.
- Its Markdown output cuts model input tokens by 93% compared with raw page HTML, according to the company's own measurement.
- JavaScript rendering, smart wait, and media parsing handle dynamic pages plus PDF and DOCX files with zero configuration.
- The codebase ships in the open, so teams can inspect and contribute to the scraping engine behind the API.
How It Works
Developers create an API key, then call the scrape, search, crawl, map, or interact endpoints from Python, Node.js, cURL, or the CLI. A scrape request returns Markdown, JSON, screenshots, or semantic text along with a scrape ID, which the interact endpoint reuses to click, type, and navigate inside a cloud browser. Teams can also run the CLI setup command, let it pick up an existing login, and give any MCP-compatible agent Firecrawl's web tools.
Use Cases
- Research teams extract papers, news articles, expert opinions, and industry data for deep, in-depth investigations.
- AI chat products pull real-time page content so assistants answer questions with current, accurate information.
- Agent builders hand autonomous agents web search, scraping, and page-control tools through MCP or the CLI.
- Sales and growth teams enrich lead profiles with contact details, funding stage, and firmographic data from across the web.
- Product teams scrape customer sites to pre-populate new accounts and speed up user onboarding.
- Data and RAG engineers feed structured, chunk-friendly Markdown into retrieval pipelines and knowledge bases.
Why Choose This Product
Firecrawl separates itself on sourcing and transparency: it licenses fair access to web content through Firecrawl Alexandria instead of relying only on raw crawling, and it publishes its benchmark methodology plus latency comparisons against Puppeteer and cURL. One endpoint set spans a single page scrape and million-page pipelines, and the Enterprise tier adds single sign-on, SCIM provisioning, a support SLA, and zero-data retention for regulated teams.
Firecrawl Pros & Cons
- Free tier with 1,000 monthly credits and no credit card
- Open source engine with 180,000+ GitHub stars
- 96% coverage on its own 1,000-URL scrape benchmark
- 93% fewer model input tokens versus raw HTML
- MCP and CLI setup in a single command
- Single sign-on and SCIM provisioning only on Enterprise
- Shared Slack support channel starts at the Growth plan
- Unused credits do not roll over below the Scale plan
Key Features
Web Search
Search the web and receive full page content from results rather than links alone, feeding fresh context into agents and RAG pipelines.
Any-Format Scrape
Convert any URL into clean Markdown, structured JSON, screenshots, or semantic text, including output shaped by a natural-language description.
Page Interact
Run cloud browser sandboxes that click, type, scroll, and navigate dynamic pages through AI prompts or code.
Crawl And Map
Discover and map every URL on a site, then batch crawl pages while sharing one credit balance across endpoints.
Agent Runs
Firecrawl Agent handles autonomous data gathering that searches, scrapes, and assembles results without custom orchestration.
Media Parsing
Parse and output content from PDF, DOCX, and other document formats alongside standard HTML pages.
Smart Wait
Detect when dynamic content finishes loading so extraction returns complete data from single-page applications.
MCP And CLI
Connect Firecrawl to any MCP client or AI coding tool in one command through the official CLI and installed skills.
Alexandria Index
Pull from a licensed, growing web index for speed or go live to the open web when the freshest data matters.
Open Source
Firecrawl publishes its scraping engine on GitHub, where the repository has passed 180,000 stars and takes community contributions.
Firecrawl Pricing
View full pricing β- 1,000 credits per month
- 500 searches or 1,000 pages scraped
- 2 concurrent requests
- Low rate limits
- No credit card required
- 5,000 credits per month
- 2,500 searches or 5,000 pages scraped
- 5 concurrent requests
- Basic support
- 1,000 extra credits per $5
- 100 requests per minute on /scrape
- 100,000 credits per month
- 50,000 searches or 100,000 pages scraped
- 25 concurrent requests
- Standard support
- 2,000 extra credits per $5
- 500 requests per minute on /scrape
- 500,000 credits per month
- 250,000 searches or 500,000 pages scraped
- 50 concurrent requests
- Priority support
- 2,500 extra credits per $5
- Shared Slack support channel
- 1,000,000 credits per month
- 500,000 searches or 1,000,000 pages scraped
- 100 concurrent requests
- Priority support
- 5,000 extra credits per $5
- One month of credit rollover
- Custom credits and concurrent requests
- Dedicated support and SLA
- Bulk discounts
- Zero-data retention
- SSO and SCIM provisioning
| Feature | Free Free | Hobby $19/mo | Standard $99/mo | Growth $399/mo | Scale $749/mo | Enterprise Custom |
|---|---|---|---|---|---|---|
| 1,000 credits per month | ||||||
| 500 searches or 1,000 pages scraped | ||||||
| 2 concurrent requests | ||||||
| Low rate limits | ||||||
| No credit card required | ||||||
| 5,000 credits per month | ||||||
| 2,500 searches or 5,000 pages scraped | ||||||
| 5 concurrent requests | ||||||
| Basic support | ||||||
| 1,000 extra credits per $5 | ||||||
| 100 requests per minute on /scrape | ||||||
| 100,000 credits per month | ||||||
| 50,000 searches or 100,000 pages scraped | ||||||
| 25 concurrent requests | ||||||
| Standard support | ||||||
| 2,000 extra credits per $5 | ||||||
| 500 requests per minute on /scrape | ||||||
| 500,000 credits per month | ||||||
| 250,000 searches or 500,000 pages scraped | ||||||
| 50 concurrent requests | ||||||
| Priority support | ||||||
| 2,500 extra credits per $5 | ||||||
| Shared Slack support channel | ||||||
| 1,000,000 credits per month | ||||||
| 500,000 searches or 1,000,000 pages scraped | ||||||
| 100 concurrent requests | ||||||
| 5,000 extra credits per $5 | ||||||
| One month of credit rollover | ||||||
| Custom credits and concurrent requests | ||||||
| Dedicated support and SLA | ||||||
| Bulk discounts | ||||||
| Zero-data retention | ||||||
| SSO and SCIM provisioning |
Pricing extracted from the product website and may change. Check the source for current details.
Frequently asked questions about Firecrawl
Is there a free Firecrawl plan?
Yes. The Free plan includes 1,000 credits every month, refreshed monthly, with no credit card required. It supports 500 searches or 1,000 pages scraped and 2 concurrent requests at low rate limits.
What counts as a credit in Firecrawl?
One credit equals one page on a basic scrape or crawl and one credit per Map call. Search costs 2 credits per 10 results, and Interact costs 2 credits per browser minute. The JSON, Question, and Highlight formats add 4 credits per page on Scrape and Crawl.
Do failed scrape requests use credits?
A scrape that returns no result is not charged. A page that responds with an error status such as 403 or 404 is still returned to you and costs 1 credit.
How much does Firecrawl cost?
Hobby costs $19 per month billed monthly, or $16 per month billed annually, for 5,000 credits. Standard runs $99 monthly ($83 annually) for 100,000 credits, Growth $399 ($333) for 500,000, and Scale $749 ($599) for 1,000,000. Enterprise pricing is custom.
Is Firecrawl open source?
Yes. Firecrawl develops its scraping engine publicly on GitHub, where the repository has passed 180,000 stars and sits among the top 100 repos. The company invites contributors to join the community.
How do I connect Firecrawl to my AI agent?
Run the CLI init command, or connect through MCP, and Firecrawl installs skills and web tools your agent can call. One command gives Cursor, Claude Code, Windsurf, and other MCP clients fresh search and scrape capability.
