gemini-mcp

Overview

This is a robust MCP server that facilitates integration between Claude and Google's Gemini AI systems. It offers an extensive collection of 37 tools spanning text processing, image generation with multi-turn editing capabilities, video creation using Veo, document analysis, and web search functionality. The server provides well-documented installation options for both Claude integration and standalone CLI usage, with extensive configuration options including tool presets to reduce context usage.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• gemini-queryprompt, model, thinkingLevel?

  Direct queries to Gemini with thinking level control

  Try:Explain quantum entanglementTry:Summarize the latest tech news
• gemini-generate-imageprompt, style?, aspectRatio?, imageSize?, useGoogleSearch?, thinkingLevel?, personGeneration?, seed?

  Generate images with Nano Banana Pro

  Try:Create a futuristic city at sunsetTry:Generate a cyberpunk landscape
• gemini-start-image-editprompt, aspectRatio?, imageSize?, useGoogleSearch?, thinkingLevel?, personGeneration?, seed?

  Start a multi-turn image editing session

  Try:Edit a cozy cabin in the mountainsTry:Start refining a product design
• gemini-continue-image-editsessionId, prompt

  Continue refining an image from an editing session

  Try:Add snow to the cabin roofTry:Change the background to a forest
• gemini-analyze-codecode, language?, focus?

  Analyze code for quality, security, or issues

  Try:Review this JavaScript functionTry:Check for security vulnerabilities in my Python code
• gemini-analyze-texttext, type?

  Analyze text for sentiment, entities, or summaries

  Try:Analyze sentiment in customer reviewsTry:Extract entities from this article
• gemini-brainstormprompt, claudeThoughts?, maxRounds?

  Collaborative brainstorming with Gemini

  Try:Brainstorm marketing strategiesTry:Collaborate on app design ideas
• gemini-summarizecontent, length?, format?

  Summarize content with different detail levels

  Try:Summarize this research paperTry:Create bullet points from this meeting transcript
• gemini-run-codeprompt, data?

  Let Gemini write and execute Python code

  Try:Calculate statistics from this dataTry:Generate a chart showing trends
• gemini-searchquery, returnCitations?

  Real-time web search with citations

  Try:Search for latest AI developmentsTry:Find information about quantum computing
• gemini-youtubeurl, question, startTime?, endTime?

  Analyze YouTube videos with timestamp clipping

  Try:Analyze this tutorial videoTry:What happens at 2:30 in this demonstration?
• gemini-analyze-documentfilePath, question, mediaResolution?

  Analyze PDFs and documents with table extraction

  Try:Summarize key findings in this reportTry:Extract tables from this research paper

Comparable tools

Installation

# Using npm (Recommended)
claude mcp add gemini -s user -- env GEMINI_API_KEY=YOUR_KEY npx -y @rlabs-inc/gemini-mcp

# Using bun
claude mcp add gemini -s user -- env GEMINI_API_KEY=YOUR_KEY bunx @rlabs-inc/gemini-mcp

npm install -g @rlabs-inc/gemini-mngcli config set api-key YOUR_KEY

FAQ

Which Gemini models are supported?

The server supports Gemini 3 Pro and Flash models for text queries, Nano Banana Pro for image generation, and Veo 2.0 for video creation.

Can I limit which tools are loaded?

Yes, you can use preset configurations (minimal, text, image, research, media, full) or specify individual tools via the GEMINI_ENABLED_TOOLS environment variable.

Compare gemini-mcp with