AI导哈

深度研究代理服务器

一个暴露Gemini深度研究代理的MCP服务器，用于全面的网络研究。

  • by bharatvansh
  • Python
  • 本地部署
  • 4 星标
  • 更新于
在 GitHub 查看 bharatvansh/gemini-deep-research-mcp

使用说明

Gemini Deep Research MCP

An MCP server that exposes Gemini's Deep Research Agent for comprehensive web research.

One-Click Install

IDE Install
Cursor
VS Code
VS Code Insiders

Note: After clicking, replace your-api-key with your Gemini API key. VS Code requires version 1.101+.

Installation Methods

Using npx (Node.js)

Requires Node.js 16+ and uv.

npx @bharatvansh/gemini-deep-research-mcp
VS Code config 
{
  "servers": {
    "gemini-deep-research": {
      "command": "npx",
      "args": ["-y", "@bharatvansh/gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Claude Desktop config 
{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "npx",
      "args": ["-y", "@bharatvansh/gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Windsurf config

Add to ~/.codeium/windsurf/mcp_config.json (macOS/Linux) or %USERPROFILE%\.codeium\windsurf\mcp_config.json (Windows):

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "npx",
      "args": ["-y", "@bharatvansh/gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Cline config 
{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "npx",
      "args": ["-y", "@bharatvansh/gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Claude Code config

Add to ~/.claude/settings.json:

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "npx",
      "args": ["-y", "@bharatvansh/gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Codex config

Add to ~/.codex/config.toml:

[mcp_servers.gemini-deep-research]
command = "npx"
args = ["-y", "@bharatvansh/gemini-deep-research-mcp"]

[mcp_servers.gemini-deep-research.env]
GEMINI_API_KEY = "your-api-key"
Cursor config

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "npx",
      "args": ["-y", "@bharatvansh/gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Antigravity config

Add to your Antigravity mcp_config.json:

{
  "gemini-deep-research": {
    "command": "npx",
    "args": ["-y", "@bharatvansh/gemini-deep-research-mcp"],
    "env": {
      "GEMINI_API_KEY": "your-api-key"
    }
  }
}

Using uvx (Python)

Requires uv.

uvx gemini-deep-research-mcp
VS Code config 
{
  "servers": {
    "gemini-deep-research": {
      "command": "uvx",
      "args": ["gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Claude Desktop config 
{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "uvx",
      "args": ["gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Windsurf config

Add to ~/.codeium/windsurf/mcp_config.json (macOS/Linux) or %USERPROFILE%\.codeium\windsurf\mcp_config.json (Windows):

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "uvx",
      "args": ["gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Cline config 
{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "uvx",
      "args": ["gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Claude Code config

Add to ~/.claude/settings.json:

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "uvx",
      "args": ["gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Codex config

Add to ~/.codex/config.toml:

[mcp_servers.gemini-deep-research]
command = "uvx"
args = ["gemini-deep-research-mcp"]

[mcp_servers.gemini-deep-research.env]
GEMINI_API_KEY = "your-api-key"
Cursor config

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "uvx",
      "args": ["gemini-deep-research-mcp"],
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Antigravity config

Add to your Antigravity mcp_config.json:

{
  "gemini-deep-research": {
    "command": "uvx",
    "args": ["gemini-deep-research-mcp"],
    "env": {
      "GEMINI_API_KEY": "your-api-key"
    }
  }
}

Using pip

pip install gemini-deep-research-mcp
VS Code config 
{
  "servers": {
    "gemini-deep-research": {
      "command": "gemini-deep-research-mcp",
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Claude Desktop config 
{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "gemini-deep-research-mcp",
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Windsurf config

Add to ~/.codeium/windsurf/mcp_config.json (macOS/Linux) or %USERPROFILE%\.codeium\windsurf\mcp_config.json (Windows):

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "gemini-deep-research-mcp",
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Cline config 
{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "gemini-deep-research-mcp",
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Claude Code config

Add to ~/.claude/settings.json:

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "gemini-deep-research-mcp",
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Codex config

Add to ~/.codex/config.toml:

[mcp_servers.gemini-deep-research]
command = "gemini-deep-research-mcp"

[mcp_servers.gemini-deep-research.env]
GEMINI_API_KEY = "your-api-key"
Cursor config

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "gemini-deep-research": {
      "command": "gemini-deep-research-mcp",
      "env": {
        "GEMINI_API_KEY": "your-api-key"
      }
    }
  }
}
Antigravity config

Add to your Antigravity mcp_config.json:

{
  "gemini-deep-research": {
    "command": "gemini-deep-research-mcp",
    "env": {
      "GEMINI_API_KEY": "your-api-key"
    }
  }
}

Antigravity

  1. Open the Agent side panel → click ... → MCP Store
  2. Search for your MCP server or click Add Custom Server
  3. Add this configuration to your mcp_config.json:
{
  "gemini-deep-research": {
    "command": "uvx",
    "args": ["gemini-deep-research-mcp"],
    "env": {
      "GEMINI_API_KEY": "your-api-key"
    }
  }
}

Prerequisites

Install uv (required for npx/uvx methods) 
# macOS/Linux
curl -LsSf https://astral.sh/uv/install.sh | sh

# Windows (PowerShell)
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Tools

1. start_deep_research

Initiates a deep, multi-step web research job in the background using Google's Deep Research Agent. Returns a job_id, which you can use to check the status of completion using check_deep_research(job_id=...).

Parameter Type Required Default Description
prompt string ✓ — Your comprehensive research question or topic
Output Description
job_id Unique tracking ID for the research job
status Initial job state (e.g. in_progress)

2. check_deep_research

Checks the status of a Deep Research job using its job_id and returns the complete report once finished.

Parameter Type Required Default Description
job_id string ✓ — The research tracking ID from start_deep_research
include_citations boolean true Include source URLs in the report
Output Description
job_id The tracking ID of the research job
status Exact current job state (in_progress, completed, failed, or cancelled)
report_text Synthesized markdown research report (when completed)
uptime Elapsed time while the job is in progress, when available
error Failure details, when the API provides them

Configuration

Variable Required Default Description
GEMINI_API_KEY ✓ — Your Gemini API key
GEMINI_DEEP_RESEARCH_AGENT deep-research-preview-04-2026 Deep Research agent to use. Set to deep-research-max-preview-04-2026 for maximum thoroughness, or deep-research-preview-04-2026 for standard speed.

Development

git clone https://github.com/bharatvansh/gemini-deep-research-mcp.git
cd gemini-deep-research-mcp
pip install -e .[dev]
pytest

License

MIT

相关 MCP