Deepsearch MCP Server

Connect Deepsearch to Claude, Cursor or any other MCP client and it stops being a tab you switch to. Gemini DeepSearch MCP is an automated research agent that leverages Google Gemini models and Google Search to perform deep, multi-step web research. It generates sophisticated queries, synthesizes information from search results. The deepsearch mcp server is what makes that connection.

What the server does

• Automated multi-step research — using Gemini models and Google Search
• FastMCP integration — for both HTTP API and stdio deployment
• Configurable effort levels — (low, medium, high) for research depth
• Citation-rich responses — with source tracking
• LangGraph-powered workflow — with state management

Available tools

The toolset is worth reading before you wire it up, because it tells you what the integration is really for:

• Low — 1 query, 1 loop, Flash model
• Medium — 3 queries, 2 loops, Flash model
• High — 5 queries, 3 loops, Pro model
• Testing — With Langsmith tracing bash GEMINI_API_KEY=AI******* LANGSMITH_API_KEY=ls******* LANGSMITH_TRACING=true make inspect
• macOS — Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
• Windows — The Windows tool exposed by this server
• Linux — The Linux tool exposed by this server

Credentials and setup notes

Configuration is passed through the environment: GEMINI_API_KEY, LANGSMITH_API_KEY. Treat anything key-shaped as a real credential — scope it to the minimum the server needs, and rotate it if it ever lands in a shared config.

• Python 3.12+ - GEMINI_API_KEY environment variable

Installation

The server ships on PyPI as install, so your MCP client can launch it on demand — there is no separate build step. Add the server block to your client's configuration, restart it, and the tools register themselves.

Where it fits

Plenty of developer tooling servers cover similar ground. The differences that matter in practice are scope of access and how much setup stands between you and a working tool call. Deepsearch's toolset — Low, Medium, High and 4 more — is a fair guide to whether it matches your workflow. It is maintained by mottysisam; worth a glance at recent repository activity before you build anything load-bearing on it.

We check each listing at SyncDev against the project's documentation before it goes live — if something here drifts out of date, it is a bug worth reporting.

Worth knowing first

• It runs with your machine's permissions. That is convenient and also the reason to think about what you point it at before you approve a tool call.
• Missing credentials fail quietly in some clients — if no tools show up, check the environment block first.
• Keep per-call confirmation enabled while you learn its behaviour; it is the cheapest safeguard you have.