Firecrawl
Firecrawl turns websites into LLM-ready data with scraping and crawling. See 2026 credit pricing, new Monitor endpoint, and top alternatives.
About Firecrawl
Firecrawl is an API that turns websites into LLM-ready data, in the chatbot and virtual assistant category, letting developers scrape, crawl, search, and interact with web pages and return clean markdown or structured content for AI applications.
What is Firecrawl (2026 update)
Firecrawl exposes web scraping, crawling, site mapping, and search as API endpoints, so developers can pull content from websites and feed it directly into LLM pipelines and AI agents without building their own scraping infrastructure. It handles JavaScript rendering, smart waiting for dynamic content, media parsing, and can extract content from PDFs, DOCX, and other document formats.
What is new in 2026
In 2026, Firecrawl launched a web-scale /monitor endpoint, an always-on search that watches the web for new pages, dedupes results, judges relevance, and pings users or agents by webhook or email, extending monitoring beyond a fixed list of known URLs to the entire web. Firecrawl also released two open-source Rust libraries for document conversion: pdf-inspector for fast PDF routing and native text extraction, and AnyDoc for one-call markdown conversion across 14 non-PDF formats, both running locally without an API key or extra system dependencies. The /search endpoint also got a major upgrade in 2026 with a new custom relevance model that returns more useful excerpts per query, improving agent search accuracy while cutting token usage, without changing the API itself.
Key features
- Scrape, Crawl, Map, and Monitor endpoints for pulling web content into markdown or structured data
- New /monitor endpoint for always-on, web-scale change detection with webhook or email alerts
- JavaScript rendering and smart wait for dynamic pages
- Document parsing for PDFs, DOCX, and other formats, including open-source Rust libraries pdf-inspector and AnyDoc
- Upgraded /search endpoint with a custom relevance model for better excerpt quality
Pricing in 2026
Firecrawl bills on a credit system, where core endpoints (Scrape, Crawl, Map, and Monitor) cost 1 credit per page, Search costs 2 credits per 10 results, Interact costs 2 credits per browser minute, and Stealth Mode costs 5 credits per page. Prices below are from the official pricing page.
| Plan | Price | What you get |
|---|---|---|
| Free | $0/month | 1,000 credits/month, 2 concurrent requests, about 1,000 pages scraped or 500 searches |
| Hobby | $16/month billed annually, or $19/month billed monthly | 5,000 credits/month, 5 concurrent requests, extra credits at 1,000 per $5 |
| Standard | $83/month billed annually (saves $198/year) | 100,000 credits/month, 25 concurrent requests, extra credits at 2,000 per $5 |
| Growth | $333/month billed annually (saves $798/year) | 500,000 credits/month, 50 concurrent requests, extra credits at 2,500 per $5 |
| Scale | $599/month billed annually (saves $1,798/year) | 1,000,000 credits/month, 100 concurrent requests, credits roll over one month, extra credits at 5,000 per $5 |
| Enterprise | Custom, contact sales | Custom credits and concurrency, dedicated support, SSO, zero-data retention, API key restrictions |
Who should use it
Firecrawl fits developers building AI agents, RAG pipelines, or research tools that need clean, structured web content rather than raw HTML. The Free and Hobby plans suit prototyping and small projects, Standard and Growth fit production apps with regular scraping volume, and Scale or Enterprise fit companies running large-scale or compliance-sensitive crawling operations.
Limitations and alternatives
Costs can climb quickly for use cases that lean heavily on Search or Stealth Mode, since those cost more credits per action than a basic page scrape, and only the Scale plan lets unused credits roll over to the next month. Enterprise features like SSO and zero-data retention are not available below the custom Enterprise tier.
Alternatives to consider in 2026 include:
- Apify, a broader web scraping and automation platform with a marketplace of pre-built scrapers
- ScrapingBee, an API-based scraping service with proxy rotation and JavaScript rendering
- Bright Data, a larger-scale web data platform aimed at enterprise data collection
Frequently asked questions
Is Firecrawl free to use?
Yes, Firecrawl has a free plan with 1,000 credits per month and 2 concurrent requests, no credit card required, enough for about 1,000 pages scraped or 500 searches.
How much does Firecrawl cost in 2026?
Paid plans start at $16/month billed annually (or $19/month billed monthly) for Hobby, rising to $83/month for Standard, $333/month for Growth, and $599/month for Scale, according to the official pricing page.
What is the /monitor endpoint?
Launched in 2026, /monitor is an always-on, web-scale search that watches for new pages across the web, dedupes and judges relevance, and alerts users or agents by webhook or email.
Do unused Firecrawl credits roll over?
Only on the Scale plan, where credits roll over for one month. Other plans do not mention credit rollover on the official pricing page.
Information last verified: September 2026.
Related Tools
