FineData

You can run the FineData MCP Server locally to enable AI agents to scrape websites while bypassing common anti-bot measures, rendering JavaScript, solving captchas, and rotating proxies. This MCP server couples with compatible clients to offer scalable web scraping through token-based usage and configurable features.

How to use

You will connect your MCP client to a local or remote MCP server by using the provided startup commands. The server accepts requests from clients to scrape pages with optional JavaScript rendering, anti-bot bypass, captcha solving, and proxy rotation. You can perform single-page scrapes, asynchronous long-running scrapes, or batch multiple URLs in one call. Use the API key you obtain from FineData to authorize requests.

How to install

Prerequisites and installation options provide flexible ways to start the MCP server using the tools you already have.

# Install uv if you haven't already
curl -LsSf https://astral.sh/uv/install.sh | sh

# Run directly with uvx
FINEDATA_API_KEY=fd_xxx uvx finedata-mcp

pip install finedata-mcp

# Run
FINEDATA_API_KEY=fd_xxx finedata-mcp

npx -y @finedata/mcp-server

Configuration and environment

Configure your MCP client or application to point at the FineData MCP server. You will supply the API key and optionally customize the API URL and timeout. The following environment variables are used by the server:

{
  "mcpServers": {
    "finedata": {
      "command": "uvx",
      "args": ["finedata-mcp"],
      "env": {
        "FINEDATA_API_KEY": "fd_your_api_key_here"
      }
    }
  }
}

{
  "mcpServers": {
    "finedata_npx": {
      "command": "npx",
      "args": ["-y", "@finedata/mcp-server"],
      "env": {
        "FINEDATA_API_KEY": "fd_your_api_key_here"
      }
    }
  }
}

Environment variables

The MCP server uses these environment variables to access your FineData account and configure behavior.

FINEDATA_API_KEY: Your FineData API key (required)
FINEDATA_API_URL: API URL (optional, default: https://api.finedata.ai)
FINEDATA_TIMEOUT: Default timeout in seconds (optional, default: 60)

Usage examples

These examples illustrate typical prompts you would send to your MCP-enabled agent.

# Basic scrape
Scrape https://example.com and show me the content

# JavaScript rendered page
Scrape https://spa-website.com with JavaScript rendering enabled

# Protected site with captcha
Scrape https://protected-site.com using residential proxy and captcha solving

# Batch scraping
Scrape these URLs: https://example.com/1, https://example.com/2, https://example.com/3

Troubleshooting

If you encounter common startup issues, ensure your environment is prepared and the API key is set correctly.

# No module named finedata_mcp
# Install uv and use uvx
curl -LsSf https://astral.sh/uv/install.sh | sh

{
  "command": "uvx",
  "args": ["finedata-mcp"]
}

Support

If you need help, reach out with your questions or issues using the contact methods below.

Available tools

scrape_url

Scrape a page with optional JS rendering, residential proxy, anti-bot bypass, and captcha solving with a per-feature token cost.

scrape_async

Submit an asynchronous scraping job for long-running requests and receive a job_id for status polling.

get_job_status

Query the status of an asynchronous scraping job by its job_id.

batch_scrape

Scrape multiple URLs in a single batch, up to 100 URLs, with optional JS rendering.

get_usage

Retrieve current API token usage for your account.