# Crawl4AI MCP

Crawl4AI’s server exposes web crawling capabilities to MCP-compatible clients for collecting readable web content.

Canonical URL: https://a2a-mcp.org/entry/crawl4ai-mcp

Language: en

Category: MCP Servers

Updated: 2026-09-14

## About

Crawl4AI’s server exposes web crawling capabilities to MCP-compatible clients for collecting readable web content.

## Key Features

- Connect an assistant to a self-hosted crawling service.
- Use Crawl4AI’s web-to-Markdown and extraction capabilities in agent workflows.

## Use Cases

Extract web content into a form an agent can use for research or data collection.

## Limitations

Use the configuration matching the deployed release. Optional model-backed extraction requires its own provider setup.

## Sources

- Project documentation: https://github.com/unclecode/crawl4ai/blob/main/deploy/docker/README.md
- Maintainer repository: https://github.com/unclecode/crawl4ai

## MCP Setup

### Prerequisites

- Docker 20.10+ with Docker Compose.
- At least 4 GB RAM available to the container.

### Client setup

#### Claude Code

Transport: `sse`

```bash
claude mcp add --transport sse c4ai-sse http://localhost:11235/mcp/sse
```

- Deploy the Crawl4AI Docker service using the repository Docker guide before connecting.

The guide also describes a WebSocket integration. Match authentication to the deployed service settings.

Setup docs: https://github.com/unclecode/crawl4ai/blob/main/deploy/docker/README.md

## Tags

- MCP
- web-extraction
- markdown

## Also known as

- Crawl4AI

## Alternatives

- Firecrawl MCP Server
- AgentQL MCP Server

## Sources

- [Project website](https://github.com/unclecode/crawl4ai/blob/main/deploy/docker/README.md)
- [Project documentation](https://github.com/unclecode/crawl4ai/blob/main/deploy/docker/README.md)
- [Maintainer repository](https://github.com/unclecode/crawl4ai)

Last checked: 2026-09-14