Snapshot
Continue to About ↓- What it does
- Firecrawl is a web data API that turns live websites into clean, LLM-ready data such as Markdown, JSON, screenshots and semantic text for AI agents and applications.
More detail
- It combines web search, scraping, crawling, site mapping, document parsing, recurring monitoring and a cloud browser sandbox for interacting with pages.
- The hosted product adds proprietary proxy and rendering infrastructure, while the core code is also available as open source for self-hosting.
- Best for
- Developer and AI product teams that need reliable, structured web context for RAG pipelines, deep-research agents, AI chat tools, lead enrichment, onboarding or price and competitive monitoring.
More detail
- Also a good fit for teams wiring web tools into MCP clients such as Cursor, Claude and Windsurf.
- Pricing
- Credit-based metered pricing in USD, shared across all endpoints, with monthly and annual rates: Free $0 for 1,000 credits
More detail
- Hobby $19/month ($16 annual) for 5,000
- Standard $99 ($83) for 100,000
- Growth $399 ($333) for 500,000
- Scale $749 ($599) for 1,000,000
- Enterprise custom.
- Credits are consumed
- Main limitation
- Teams that need long-running, persistent authenticated browser sessions, since Interact is not designed for that, or that require browser automation fully offline on self-hosted installs, because Interact is cloud-only.
More detail
- The free tier (1,000 credits, 2 concurrent requests, low rate limits) is also too thin for production-scale work.
- Record
- Living and monitored
- Reviewed
- Evidence
- Public web research
- Machine access
- Structured profile ↗
About
- Firecrawl is a web data API that turns live websites into clean, LLM-ready data such as Markdown, JSON, screenshots and semantic text for AI agents and applications.
- It combines web search, scraping, crawling, site mapping, document parsing, recurring monitoring and a cloud browser sandbox for interacting with pages.
- The hosted product adds proprietary proxy and rendering infrastructure, while the core code is also available as open source for self-hosting.
What it helps with
- Feeding LLMs and AI agents clean, real-time web context without assembling proxies, renderers and post-processing yourse
- Getting full page content, not just links, from web search results in a single call
- Extracting structured JSON or screenshots from JavaScript-heavy and dynamic pages
- Crawling and mapping whole sites and parsing PDFs or DOCX documents into usable text
- Automating clicks, form fills and navigation on pages that require interaction to reach data
- Running recurring page or search checks with webhook, email or Slack alerts for monitoring
Capabilities
A practical view of what is available and where limitations apply.
| Offering | Available | Detail |
|---|---|---|
| Search | Yes | Queries the live web and can return full page content with results; costs 2 credits per 10 results. |
| Scrape | Yes | Converts a URL to Markdown, JSON, screenshots or semantic text; 1 credit per page, with +4 credits per page for JSON, question or highlight formats. |
| Crawl and Map | Yes | Site traversal and structure discovery, each billed at 1 credit per page. |
| Interact (cloud browser) | Limited | Click, fill, navigate and log in via AI prompts or code at 2 credits per browser minute; cloud-only and not designed for long-running persistent sessions. |
| Document parsing | Yes | Parses PDFs, DOCX and other documents; billed per PDF page. |
| Monitoring | Yes | Recurring checks on pages or searches with webhook, email or Slack alerts; 1 credit per page per check, or 7 in deterministic-extraction mode. |
| MCP server and CLI | Yes | Official MCP server for Cursor, Claude, Windsurf and other clients, plus a CLI and agent skills. |
| Open source / self-hosting | Yes | Core is open source and self-hostable; hosted-only extras include Fire-engine, Interact and the analytics dashboard. |
| Free tier | Limited | 1,000 credits per month with 2 concurrent requests and low rate limits; no credit card required. |
| Independent review evidence | Unknown | No credible independent review sources were found in the supplied research; testimonials cited are vendor-published. |
Bot trust
How confidently AI systems can verify this profile from public evidence.
48of 100
ChatGPT—/100
Claude48/100
What bots could verify
- • Corporate identity is clearly documented with verifiable founders, founding year, Y Combinator backing, and notable investors.
- • Appropriately distinguishes vendor-reported metrics and customer logos from independently corroborated evidence.
- • Highlights material operational, legal, and compliance risks inherent to web data extraction.
What weakens trust
- • Pricing transparency is completely absent, lacking tier costs, free tier limits, credit structures, and rate limits.
- • Zero independent customer evidence or third-party platform reviews (e.g., G2, Capterra, or Trustpilot) were retrieved.
- • No compliance or security verifications were located, leaving SOC 2 certification, data retention, and privacy terms unknown.
Evidence
Where this record came from, when it was reviewed, and how confidently it can be used.
- Verification
- Scribble researched these facts from the linked public sources.
- The owner has not manually confirmed every field.
- Last reviewed
- Primary source
- firecrawl.dev
Technical profile details
Record ID
8de56720-be58-4e68-abfb-0d7cf72f9e67Canonical URLhttps://profiles.scribble.network/listing/firecrawl-8de567