Gemini Image Generator MCP Server

What is Gemini Image Generator MCP Server?

This MCP server enables AI assistants to generate high-quality images from text prompts and transform existing images using Google's Gemini model via the MCP protocol. It handles prompt engineering, text-to-image conversion, intelligent filename generation, and local image storage. Aimed at users of MCP-compatible clients such as Claude Desktop, Cursor, or other hosts.

How to use Gemini Image Generator MCP Server?

Install via Smithery (npx -y @smithery/cli install @qhdrl12/mcp-server-gemini-image-gen --client claude), or manually clone the repo, create a Python 3.11+ virtual environment, and install dependencies. Configure a Gemini API key (via Google AI Studio) and an optional image output path. Add the server to your MCP client’s configuration (e.g., claude_desktop_config.json) with the command, args, and env keys. Invoke the three available MCP tools: generate_image_from_text, transform_image_from_encoded, or transform_image_from_file.

Key features of Gemini Image Generator MCP Server

• Text-to-image generation using Gemini 2.0 Flash
• Image-to-image transformation based on text prompts
• Support for both file-based and base64-encoded images
• Automatic intelligent filename generation from prompts
• Automatic translation of non-English prompts
• Local image storage with configurable output path

Use cases of Gemini Image Generator MCP Server

• Generate an image from a natural-language description (e.g., "a sunset over mountains")
• Transform an existing image by adding or modifying elements (e.g., "add snow to this landscape")
• Edit a photo by providing a base64-encoded image and a transformation prompt
• Use the server as a creative assistant for generating illustrations, concept art, or visual ideas

FAQ from Gemini Image Generator MCP Server

What does this server do vs. other image generation servers?

It provides text-to-image and image-to-image transformation specifically via Google’s Gemini Flash model, with dual return (image bytes and saved file path). It also offers automatic non-English prompt translation and strict text exclusion from generated images.

What runtime dependencies are required?

Python 3.11 or higher, a Google AI API key (Gemini), and an MCP-compatible host application (e.g., Claude Desktop, Cursor).

Where are generated images stored?

Images are saved to a local directory specified by the OUTPUT_IMAGE_PATH environment variable. The server also returns raw image data (bytes) for direct use.

Are there known limitations or issues?

When used with Claude Desktop Host, transform_image_from_encoded may perform slowly due to large base64 data transfer, and path resolution problems can occur with returned file paths. Using alternative MCP clients or the transform_image_from_file method is recommended.

How does authentication work?

Authentication requires a Google AI API key, set via the GEMINI_API_KEY environment variable in the MCP client configuration or a .env file. The API key provides a certain free quota per month.