Skip to main content
Official CAMEL docs: docs.camel-ai.org/key_modules/loadersFirecrawl ships as a CAMEL loader, turning any site into LLM-ready markdown for your agents

Installation

The web-tools extra installs the Firecrawl SDK version the loader supports. A plain pip install camel-ai doesn’t include it, and the current CAMEL release fails with firecrawl-py 4.x, so don’t upgrade firecrawl-py separately in the same environment. Set your Firecrawl key so the loader picks it up automatically:

Usage

With Camel AI and Firecrawl you can quickly build multi-agent systems that use data from the web.

Using Firecrawl to Gather an Entire Website

crawl returns once the job finishes. Check status, then read the markdown from data.

Using Firecrawl to Gather a Single Page

Mapping a Site and Extracting Structured Data

map_site lists a site’s URLs, and structured_scrape returns fields that match a Pydantic model.
Firecrawl() reads FIRECRAWL_API_KEY from the environment. Pass api_key=... to override it, or api_url=... to point at a self-hosted instance.

Resources