AI Studio MCP Server

A Model Context Protocol (MCP) server that integrates with Google AI Studio / Gemini API, providing content generation capabilities with support for files, conversation history, and system prompts.

Installation and Usage

Prerequisites

Node.js 20.0.0 or higher
Google AI Studio API key

Using npx (Recommended)

GEMINI_API_KEY=your_api_key npx -y aistudio-mcp-server

Local Installation

npm install -g aistudio-mcp-server
GEMINI_API_KEY=your_api_key aistudio-mcp-server

Configuration

Set your Google AI Studio API key as an environment variable:

export GEMINI_API_KEY=your_api_key_here

Optional Configuration

GEMINI_MODEL: Gemini model to use (default: gemini-2.5-flash)
GEMINI_TIMEOUT: Request timeout in milliseconds (default: 300000 = 5 minutes)
GEMINI_MAX_OUTPUT_TOKENS: Maximum output tokens (default: 8192)
GEMINI_MAX_FILES: Maximum number of files per request (default: 10)
GEMINI_MAX_TOTAL_FILE_SIZE: Maximum total file size in MB (default: 50)
GEMINI_TEMPERATURE: Temperature for generation (0-2, default: 0.2)

Example:

export GEMINI_API_KEY=your_api_key_here
export GEMINI_MODEL=gemini-2.5-flash
export GEMINI_TIMEOUT=600000  # 10 minutes
export GEMINI_MAX_OUTPUT_TOKENS=16384  # More output tokens
export GEMINI_MAX_FILES=5  # Limit to 5 files per request
export GEMINI_MAX_TOTAL_FILE_SIZE=100  # 100MB limit
export GEMINI_TEMPERATURE=0.7  # More creative responses

Available Tools

generate_content

Generates content using Gemini with comprehensive support for files, conversation history, and system prompts. Supports various file types including images, PDFs, Office documents, and text files.

Parameters:

user_prompt (string, required): User prompt for generation
system_prompt (string, optional): System prompt to guide AI behavior
files (array, optional): Array of files to include in generation
- Each file object must have either path or content
- path (string): Path to file
- content (string): Base64 encoded file content
- type (string, optional): MIME type (auto-detected from file extension)
model (string, optional): Gemini model to use (default: gemini-2.5-flash)
temperature (number, optional): Temperature for generation (0-2, default: 0.2). Lower values produce more focused responses, higher values more creative ones

Supported file types (Gemini 2.5 models):

Images: JPG, JPEG, PNG, GIF, WebP, SVG, BMP, TIFF
Video: MP4, AVI, MOV, WEBM, FLV, MPG, WMV (up to 10 files per request)
Audio: MP3, WAV, AIFF, AAC, OGG, FLAC (up to 15MB per file)
Documents: PDF (treated as images, one page = one image)
Text: TXT, MD, JSON, XML, CSV, HTML

File limitations:

Maximum file size: 15MB per audio/video/document file
Maximum total request size: 20MB (2GB when using Cloud Storage)
Video files: Up to 10 per request
PDF files follow image pricing (one page = one image)

Basic example:

{
  "user_prompt": "Analyze this image and describe what you see",
  "files": [
    {
      "path": "/path/to/image.jpg"
    }
  ]
}

PDF to Markdown conversion:

{
  "user_prompt": "Convert this PDF to well-formatted Markdown, preserving structure and formatting. Return only the Markdown content.",
  "files": [
    {
      "path": "/path/to/document.pdf"
    }
  ]
}

With system prompt:

{
  "system_prompt": "You are a helpful document analyst specialized in technical documentation",
  "user_prompt": "Please provide a detailed explanation of the authentication methods shown in this document",
  "files": [
    {"path": "/api-docs.pdf"}
  ]
}

Multiple files example:

{
  "user_prompt": "Compare these documents and images",
  "files": [
    {"path": "/document.pdf"},
    {"path": "/chart.png"},
    {"content": "base64encodedcontent", "type": "image/jpeg"}
  ]
}

Common Use Cases

PDF to Markdown Conversion

To convert PDF files to Markdown format, use the generate_content tool with an appropriate prompt:

{
  "user_prompt": "Convert this PDF to well-formatted Markdown, preserving structure, headings, lists, and formatting. Include table of contents if the document has sections.",
  "files": [
    {
      "path": "/path/to/document.pdf"
    }
  ]
}

Image Analysis

Analyze images, charts, diagrams, or photos with detailed descriptions:

{
  "system_prompt": "You are an expert image analyst. Provide detailed, accurate descriptions of visual content.",
  "user_prompt": "Analyze this image and describe what you see. Include details about objects, people, text, colors, and composition.",
  "files": [
    {
      "path": "/path/to/image.jpg"
    }
  ]
}

For screenshots or technical diagrams:

{
  "user_prompt": "Describe this system architecture diagram. Explain the components and their relationships.",
  "files": [
    {
      "path": "/architecture-diagram.png"
    }
  ]
}

Audio Transcription

Generate transcripts from audio files:

{
  "system_prompt": "You are a professional transcription service. Provide accurate, well-formatted transcripts.",
  "user_prompt": "Please transcribe this audio file. Include speaker identification if multiple speakers are present, and format it with proper punctuation and paragraphs.",
  "files": [
    {
      "path": "/meeting-recording.mp3"
    }
  ]
}

For interview or meeting transcripts:

{
  "user_prompt": "Transcribe this interview and provide a summary of key points discussed.",
  "files": [
    {
      "path": "/interview.wav"
    }
  ]
}

MCP Client Configuration

Add this server to your MCP client configuration:

{
  "mcpServers": {
    "aistudio": {
      "command": "npx",
      "args": ["-y", "aistudio-mcp-server"],
      "env": {
        "GEMINI_API_KEY": "your_api_key_here",
        "GEMINI_MODEL": "gemini-2.5-flash",
        "GEMINI_TIMEOUT": "600000",
        "GEMINI_MAX_OUTPUT_TOKENS": "16384",
        "GEMINI_MAX_FILES": "10",
        "GEMINI_MAX_TOTAL_FILE_SIZE": "50",
        "GEMINI_TEMPERATURE": "0.2"
      }
    }
  }
}

Development

Setup

Make sure you have Node.js 20.0.0 or higher installed.

npm install
npm run build

Running locally

GEMINI_API_KEY=your_api_key npm run dev

License

MIT

AI Studio MCP Server