MCP.so
Sign In

Kagi MCP server

@kagisearch

About Kagi MCP server

The Official Model Context Protocol (MCP) server for Kagi Search & other tools.

Basic information

Category

Search

License

MIT

Runtime

python

Transports

stdio

Publisher

kagisearch

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "kagimcp": {
      "command": "npx",
      "args": [
        "-y",
        "@smithery/cli",
        "install",
        "kagimcp",
        "--client",
        "claude"
      ]
    }
  }
}

Tools

2

web, news, videos, podcasts, and image search with optional page extracts, filters, and Kagi lenses.

fetch a page's full content as markdown.

Overview

What is Kagi MCP server?

The Kagi MCP server is an MCP‑compatible server backed by the Kagi API. It exposes web, news, video, podcast, and image search tools, as well as a page‑extraction tool, enabling AI assistants to search the web and retrieve full page content as markdown.

How to use Kagi MCP server?

Use the hosted server at https://mcp.kagi.com/mcp with any HTTP‑capable MCP client, authenticating with your Kagi API key via a Bearer token. Alternatively, install the server locally using uvx kagimcp after setting the KAGI_API_KEY environment variable. Configuration options include search/extract timeouts, retry count, and hidden search parameters.

Key features of Kagi MCP server

  • Web, news, video, podcast, and image search with filters and lenses.
  • Full page extraction returning content as markdown.
  • Hosted HTTP server available at https://mcp.kagi.com/mcp — no install required.
  • Self‑hostable via the provided Dockerfile, supporting multi‑tenant HTTP mode.
  • Configurable timeouts, retries, and hidden parameters for LLM‑facing schemas.
  • Supports multiple MCP clients: Claude Desktop, Claude Code, Codex CLI, Smithery, Kiro, OpenCode.

Use cases of Kagi MCP server

  • AI assistants performing live web searches for current events or research.
  • Extracting full article or documentation content from a URL for further processing.
  • Searching across multiple media types (news, videos, images) with fine‑grained filters.

FAQ from Kagi MCP server

Do I need a Kagi account and API key?

Yes. A Kagi API key is required and must be set as the KAGI_API_KEY environment variable or passed via the Authorization: Bearer <key> HTTP header.

What are the runtime dependencies?

The recommended install uses uv. For local setups, uv must be installed. The server runs as a Python package (kagimcp) via uvx.

Does the server support OAuth2 authentication?

Not yet. OAuth2 is on the roadmap; for now you must use a Kagi API key with Bearer authentication.

Where do search queries and extracted data go?

All requests are sent to the Kagi API. The server does not store data locally; results are returned directly to the MCP client.

Can I control timeouts or retries?

Yes. Set KAGI_SEARCH_TIMEOUT (default 10s), KAGI_EXTRACT_TIMEOUT (default 30s), and KAGI_MAX_RETRIES (default 2) environment variables.

Comments

More Search MCP servers