ai.smithery/Artin0123-gemini-image-mcp-server
Analyze images and videos with Gemini to get fast, reliable visual insights. Handle content from U…
Open source Repository Open in the app JSON README (API)
About
Analyze images and videos with Gemini to get fast, reliable visual insights. Handle content from U…
Details
- Kind
- MCP servers
- Topic
- AI, RAG & memory
- Publisher
- ai.smithery
- Origin
- official
- Category
- ferramentas
- Transport
- http
- Version
- 1.4.3
- Stars
- 2
- Last push
- 2025-10-08T07:09:41Z
- Repository state
- ativo
- Language
- TypeScript
- License
- MIT
- Added
- 2026-08-29 03:00:38
- Updated
- 2026-08-29 03:00:38
- Origin id
ai.smithery/Artin0123-gemini-image-mcp-server
README
# image-mcp-server-gemini
[](https://smithery.ai/server/@Artin0123/gemini-image-mcp-server)
> This is remote server, use [local version](https://github.com/Artin0123/gemini-vision-mcp/tree/local) for local images and videos.
## Features
- Analyze one or more image URLs with a single tool call.
- Analyze YouTube videos without downloading files locally.
- Supply an API key and optionally override the Gemini model via environment variables.
- **File size limit**: Images are limited to 16 MB to ensure fast processing.
- **YouTube videos**: No size limit as they are streamed directly by Gemini API.
## Installation
### Installing via Smithery
Install the server in Claude Desktop:
```bash
npx -y @smithery/cli install @Artin0123/gemini-image-mcp-server --client claude
```
### Manual Installation
```bash
# Clone the repository
git clone https://github.com/Artin0123/gemini-vision-mcp.git
cd gemini-vision-mcp
# Install dependencies
npm install
# Compile TypeScript to dist/
npm run build
```
## Configuration
Create a Gemini API key in [Google AI Studio](https://aistudio.google.com/app/apikey) and provide `GEMINI_API_KEY` to the server.
```json
{
"mcpServers": {
"gemini-media": {
"command": "node",
"args": ["/absolute/path/to/gemini-vision-mcp/dist/index.js"],
"env": {
"GEMINI_API_KEY": "your_api_key_here",
"GEMINI_MODEL": "models/gemini-flash-lite-latest"
}
}
}
}
```
If no key is supplied, the server can still start (handy for automated scans), but any tool invocation will return a configuration error until a valid API key is configured.
### Model override
The server defaults to `models/gemini-flash-lite-latest`. Override it by either:
> Setting the `GEMINI_MODEL` environment variable, or Providing `modelName` in the Smithery/SDK configuration schema.
## Available tools
- `analyze_image`: Analyze one or more image URLs. **Maximum file size: 16 MB per image.**
- `analyze_youtube_video`: Analyze a YouTube video from URL. No size limit.
Image URLs are downloaded and processed with a 16 MB size limit to ensure fast response times. Files exceeding this limit will result in an error message indicating the actual file size.
YouTube videos are streamed directly by Gemini API without downloading, so there is no size restriction.
### Prompt examples
```
Please analyze this product photo: https://teimg-bgr.pages.dev/file/mvYT6KeF.webp
```
```
Extract the main talking points from this clip: https://www.youtube.com/watch?v=dQw4w9WgXcQ
```
## Development
```bash
npm install
npm test
npm run build
```
The test suite exercises URL forwarding, MIME handling, and configuration fallbacks.
## License
MIT