Skip to content
Sign in
  • Browse all toolsFilter by category and pricing
  • Browse categoriesWriting, code, image, audio and more
  • Compare toolsPut up to three side by side
  • Read the blogGuides and comparisons

Firecrawl

An open-source API for web search, scraping, browser interaction, and structured data extraction for AI agents.

Firecrawl is a web data API that searches, scrapes, and interacts with websites to produce Markdown, screenshots, and structured JSON. Firecrawl serves developers building AI agents and applications, with access through an open-source repository or a hosted service that requires an API key.

Updated 9 Nov 20266 sections

Firecrawl website screenshot

What is Firecrawl?

Firecrawl is an open-source context API for AI agents that searches, scrapes, and interacts with the web. Firecrawl serves developers who need website content in formats suitable for AI applications, including Markdown and structured JSON. The project is published on GitHub and is also available as a hosted service.

Firecrawl combines content retrieval with browser interaction and automated data gathering. The repository describes support for JavaScript-heavy pages, rotating proxies, orchestration, and rate limits. Beyond web pages, Firecrawl parses PDFs, DOCX files, and other documents into Markdown or structured JSON, bringing document content into the same extraction workflow.

What Firecrawl can do

Firecrawl provides endpoints for tasks ranging from retrieving a known URL to gathering information from a natural-language request. Search returns full page content from results, while Scrape converts individual pages into selected output formats. Additional endpoints cover website discovery, crawling, document parsing, and notifications when website content changes.

  • Firecrawl searches the web and retrieves full page content from results.
  • Firecrawl converts URLs into Markdown, HTML, screenshots, or structured JSON.
  • Firecrawl interacts with pages through AI prompts or code.
  • Firecrawl parses PDFs, DOCX files, and other documents.
  • Firecrawl discovers website URLs and crawls website pages.
  • Firecrawl scrapes multiple URLs asynchronously.
  • Firecrawl gathers information from prompts and supports schema-based output.
  • Firecrawl notifies users when website content changes.

How Firecrawl works

Firecrawl hosted access starts with signing up at firecrawl.dev for an API key. The repository offers a playground for testing and examples using Python, Node.js, cURL, and the CLI. A search request takes a query, while a scrape request takes a URL and returns extracted content.

Firecrawl browser interaction begins by scraping a page and obtaining its scrape identifier. Subsequent interaction requests use that identifier with prompts or code. Crawl requests return a job identifier for checking progress, and the SDKs handle polling automatically. The Agent endpoint accepts a task description, optional URLs, and an optional output schema.

Who Firecrawl is for

Firecrawl is aimed at developers building AI agents and applications that need web context. Firecrawl connects to MCP-compatible clients and provides an agent skill installation path. The repository names Claude Code, Antigravity, and OpenCode as compatible tools, alongside examples for research, website extraction, and browser-based information gathering.

  • Firecrawl supplies current web content to AI agents.
  • Firecrawl extracts website information into structured records.
  • Firecrawl gathers research data without requiring URLs upfront.
  • Firecrawl converts documents into Markdown for downstream processing.
  • Firecrawl discovers relevant pages within a website.

Firecrawl access and licensing

Firecrawl is available as an open-source project and as a hosted service. The repository lists an AGPL-3.0 license and includes self-hosting documentation and a Docker Compose file. Hosted users obtain an API key through firecrawl.dev. The supplied page does not specify hosted subscription prices, free allowances, or plan limits.

Firecrawl strengths and limitations

Firecrawl brings search, scraping, interaction, and document parsing into a shared API workflow. Multiple output formats and SDK-managed crawl polling support application development, while MCP and agent skills provide additional connection options. The documented setup still involves API credentials, commands, or client configuration rather than only a graphical interface.

  • Firecrawl hosted API examples require an API key.
  • Firecrawl agent skill installation requires restarting the agent.
  • Firecrawl MCP connections require client configuration.
  • Firecrawl Agent effort levels change reasoning budget rather than model selection.

Firecrawl FAQ

What is Firecrawl?

Firecrawl is an open-source web data API for AI agents and applications. Firecrawl searches the web, scrapes pages, interacts with websites, and returns content as Markdown, HTML, screenshots, or structured JSON. Firecrawl is also available as a hosted service.

Is Firecrawl open source?

Firecrawl is an open-source project with an AGPL-3.0 license listed in the repository. Firecrawl also offers a hosted service, and the repository includes self-hosting documentation. The supplied page does not state hosted pricing or free usage allowances.

Does Firecrawl work with MCP clients and coding agents?

Firecrawl supports MCP-compatible clients and offers an agent skill installation command. Firecrawl names Claude Code, Antigravity, and OpenCode as compatible tools. The MCP configuration uses a Firecrawl API key, and the skill instructions require restarting the agent after installation.

What formats can Firecrawl return?

Firecrawl can return scraped website content as Markdown, HTML, screenshots, or structured JSON. Firecrawl also converts PDFs, DOCX files, and other documents into Markdown or structured JSON, while the Agent endpoint supports a supplied schema for structured results.

Can Firecrawl gather data without a list of URLs?

Firecrawl provides an Agent endpoint that accepts a description of the information needed, then searches, navigates, and retrieves results without requiring URLs upfront. Firecrawl also accepts optional URLs to focus the task and supports effort settings for different reasoning budgets.

How do developers start using Firecrawl?

Firecrawl directs hosted users to sign up at firecrawl.dev for an API key and offers a playground for testing. Firecrawl provides Python, Node.js, cURL, and CLI examples, alongside MCP configuration and agent skill installation instructions.

From the community

No reviews yet

Be the first to review this tool. What you write helps the next person choose.

More tools like Firecrawl

Crawl4AI

www.github.com

Open-source web crawling and extraction for AI, with Python, Docker, CLI and hosted API access.

Freemium

Olostep Web Data API

www.olostep.com

Web scraping, crawling, search and monitoring API for engineering teams building AI applications.

Freemium

Apify

www.apify.com

A marketplace of web scraping and automation tools for real-time data, apps, and AI agents.

Freemium

PaddleOCR

www.github.com

Open-source OCR and document parsing that converts PDFs and images into structured Markdown and JSON for AI workflows.

Free