Python SDK
Typed Python clients for web research and data pipelines.
Use Geonode MCP from your editor or chat. Start with a page extraction, then add an API key for the full tool set.
Use a supported assistant or IDE with an MCP connection.
Anonymous mode offers limited extract. Add a key for full access.
For JSON-based clients such as Cursor. Follow your client’s setup guide, then ask it to extract a public page.
{
"mcpServers": {
"geonode": {
"url": "https://mcp.geonode.com/mcp"
}
}
}Keep the same endpoint when moving from a trial call to authenticated workflows.
Anonymous limits apply per network address.
extract only
5 requests / minute
1 concurrent request
100 MB / dayAdd your API key for more tools and plan limits.
"headers": {
"X-Api-Key": "YOUR_API_KEY"
}Follow the guide for Cursor, Claude Code, VS Code and other supported clients.
https://mcp.geonode.com/mcpGet the full setup guide, authentication details and technical reference.
Extract webpage content from a compatible AI assistant or coding tool.
Try basic synchronous page extraction without an API key, within anonymous limits.
Use an API key for URL discovery, crawls, web search and batch jobs.
Access JavaScript rendering and country targeting with authenticated requests.
Monitor long-running extraction and crawl jobs from your MCP client.
Extract https://example.com as Markdown.
Summarize the main points and keep the source URL.Compatible remote MCP client
API key recommended for the full toolset. Keyless synchronous extraction is also documented: 5 requests/minute, 1 concurrent request and 100 MB/day; no JavaScript rendering or country targeting.
Save the client configuration and reconnect.
This is expected without a recognized API key.
Check X-Api-Key in the client configuration.
Wait for limits to reset or add an API key.
Verify the connection with a small extract.
Give the assistant a URL and clear output.
Review arguments and returned content in your client.
Add a key to the same MCP endpoint.
Typed Python clients for web research and data pipelines.
Web research and content collection for LangChain agents.
Try a page without a key, then add full access when needed.