Firecrawl
4.5
Firecrawl fetches web data in a form models can read, through scrape, crawl, map, search and agent endpoints, metered in credits at roughly one credit per page. Free gives 1,000 credits a month with no card required; Hobby is $19 for 5,000, Standard $99 for 100,000, Growth $399 for 500,000 and Scale $749 for one million.
- Commercial use OK
- Free tier
- Pricing
- $19/mo
- Free tier
- Free tier
- Commercial use
- Permitted
- Japanese UI
- Not confirmed
- Enterprise contract
- Not confirmed
What it is good for
- Feeding web content into RAG pipelines
- Agents that need to read pages rather than render them
- Crawling documentation sites for indexing
- Replacing a homemade scraper that keeps breaking
Strengths and trade-offs
Strengths
- Output is already shaped for model consumption, not raw HTML
- Credit pricing maps directly to pages, which is easy to forecast
- Free tier of 1,000 credits a month is usable for small jobs
- Handles JavaScript-rendered pages
Trade-offs
- Credit consumption rises above one per page for heavier operations
- Respecting robots.txt and site terms remains your responsibility
- Large crawls get expensive quickly
Editors' review
The cleanup is the value
Fetching HTML is trivial. Turning a page full of navigation, cookie banners and ads into the article text is the part that consumes a week and then breaks when the site redesigns.
At 1,000 free credits a month, a documentation site can be indexed without paying anything. Large-scale crawling is where the cost arrives.
Verdict
Good value for RAG ingestion. Model the credit cost before a large crawl.
Compiled from official documentation and public sources, with our own editorial judgement. We do not publish figures we cannot source.