Raycast AI
ProductivityAI assistance, models, and automations from Raycast
Multi-model interfaceAI Agent
Web search
Firecrawl
Context API for AI agents to search, scrape, and interact with the live web.
Firecrawl product overviewExplore the image ↓
What you can do with it
Overview and audience
Firecrawl turns websites and documents into clean, machine-usable information for AI applications. Its API can search the web, scrape individual URLs, crawl sites, map available pages, and return Markdown, HTML, metadata, screenshots, or structured JSON.
The hosted service handles JavaScript rendering, dynamic content, proxies, and interactive browser workflows. Firecrawl also provides SDKs, a CLI, an MCP server, and an open-source version, making it suitable for applications that need web context rather than a conventional browser interface.
Who it suits
Developers and teams building AI agents, RAG pipelines, research tools, chat applications, lead-enrichment workflows, content systems, and other software that needs live web data.
01
Build a research workflow that searches for sources, scrapes relevant pages, and supplies clean Markdown to an AI system for synthesis.
02
Provide an assistant with current website content by combining search or crawling with retrieval, structured extraction, and response generation.
03
Collect public company, product, or contact details from selected sites and normalize them into fields for a CRM, catalog, or internal database.
A closer look at the product
Official product image. Click to inspect the details.
Product preview
Key features, practical tasks
Key features and use cases
Web search
Search the web in one API call and receive relevant results with full-page content, reducing the need to separately discover URLs and scrape each result.
URL scraping
Convert individual webpages into clean Markdown, HTML, metadata, screenshots, or structured data generated from a schema you provide.
Interactive browsing
Continue from a scraped page by clicking, typing, scrolling, waiting, pressing controls, filling forms, or navigating multi-step workflows through prompts or code.
Site crawling and mapping
Map URLs across a website or crawl linked pages recursively, with controls for depth, page limits, and path filters.
Dynamic-page handling
Render JavaScript-driven sites and intelligently wait for content to load, helping extract information from single-page applications and dynamically populated pages.
Structured extraction
Pass a JSON schema to return fields such as product listings, pricing tables, or contact information in the shape your application expects.
Document parsing
Parse PDFs, DOCX, XLSX, HTML, and other supported documents into Markdown or structured JSON for downstream search and AI processing.
Agent integrations
Connect AI agents through the MCP server, CLI, or agent skills, with official SDKs and direct REST access for custom integrations.
Research assistant
Build a research workflow that searches for sources, scrapes relevant pages, and supplies clean Markdown to an AI system for synthesis.
Web-grounded chat
Provide an assistant with current website content by combining search or crawling with retrieval, structured extraction, and response generation.
Lead and catalog enrichment
Collect public company, product, or contact details from selected sites and normalize them into fields for a CRM, catalog, or internal database.
Get started, step by step
- 01
Use the browser Playground, REST API, an SDK, CLI, MCP server, or self-hosted open-source deployment.
- 02
Send a website URL to scrape, a query to search, or a starting URL to map or crawl.
- 03
Request Markdown, HTML, metadata, screenshots, links, or structured JSON with a schema.
- 04
Create an interaction from a scrape and describe actions such as clicking, searching, filling, or navigating.
- 05
Track credits, rate limits, concurrency, and pay-as-you-go settings before moving a workflow into larger-scale use.
Getting started
Getting started
1. Choose an access route
Use the browser Playground, REST API, an SDK, CLI, MCP server, or self-hosted open-source deployment.
2. Start with a URL or query
Send a website URL to scrape, a query to search, or a starting URL to map or crawl.
3. Select an output
Request Markdown, HTML, metadata, screenshots, links, or structured JSON with a schema.
4. Add interaction when needed
Create an interaction from a scrape and describe actions such as clicking, searching, filling, or navigating.
5. Monitor usage
Track credits, rate limits, concurrency, and pay-as-you-go settings before moving a workflow into larger-scale use.
Find the right plan for you
Plans and access
Explore official plans ↗Choose your workflow
Plans and access
A free tier provides 1,000 credits monthly without a card. Paid options include Hobby, Standard, Growth, Scale, and custom Enterprise pricing; paid plans support optional pay-as-you-go credits.
Free
$0 Monthly 1,000 credits per month, no card required, with low rate limits and two concurrent requests.
Hobby
$19 monthly or $16 monthly billed annually Monthly or annual 5,000 credits per month for side projects and small tools, with five concurrent requests and basic support.
Enterprise
Custom Custom Custom credits and concurrency with dedicated support, SLA, SSO, advanced security options, and custom terms.
A few things worth knowing
Frequently asked questions
What is Firecrawl?
Firecrawl is a context API that searches, scrapes, crawls, and interacts with the live web, returning clean content for AI agents and applications.
Does Firecrawl support JavaScript websites?
Yes. The service renders JavaScript and handles dynamically loaded content without requiring separate rendering configuration.
What formats can it return?
Supported outputs include Markdown, HTML, structured JSON, screenshots, links, and page metadata.
Can it operate webpages?
Yes. Interact can click, type, scroll, wait, press controls, fill forms, navigate, and extract information from multi-step pages.
Is Firecrawl open source?
Yes. Firecrawl provides an open-source version alongside its hosted service, whose infrastructure includes managed rendering and proxy capabilities.
How are credits charged?
Basic scrape, crawl, and map requests cost 1 credit per page; search costs 2 credits per 10 results, and interact costs 2 credits per browser minute.
The details behind your decision
Explore platforms, inputs and outputs, licensing, and access requirements.
Platforms, inputs and outputs
Platforms, inputs and outputs
- Platform
- REST API, Python, Node.js, Go, Rust, Java, Elixir, CLI, MCP-compatible AI tools, and self-hosted open-source deployment.
- Input
- Search queries, website URLs, crawl starting points, interaction prompts, JSON schemas, and local PDFs, DOCX, XLSX, or HTML files.
- Output
- Clean Markdown, HTML, structured JSON, screenshots, links, metadata, search results, crawl or map results, extracted document content, and interaction outputs.
Credit consumption varies by endpoint and output format: basic scrape, crawl, or map uses 1 credit per page; search uses 2 credits per 10 results; interact uses 2 credits per browser minute. JSON, Question, and Highlight formats add credits. Free and most paid plans have rate, concurrency, and queue limits. Hobby, Standard, and Growth credits do not roll over. Agent is described as preview with dynamic pricing.
Sources, documentation and help
Your experience can help someone choose
COMMUNITY
Sign in to rate and comment
Tried it? Share your experience
Loading experiences…
Manage my comments
Community and moderation
Saved tools are private. Approved comments are public; edits return to moderation. Comment counts include approved comments and replies. Each person has one rating. New ratings are automatically approved; withdrawn, unapproved or invalid ratings do not count. Share real experiences and avoid spam, private information or personal attacks.
For moderation appeals or data requests: support@findgoodai.com
NEXT STEP
Explore Firecrawl ↗Explore Firecrawl
Bring a real task and see how it fits the way you work.
Jump to a section
User experiences