media-mcp

🎨 media-mcp-server

MCP server for image & video processing. No API keys. No config. Just tools.

Give your AI assistant the power to resize, convert, compress, crop, filter, and analyze images and videos — all through Model Context Protocol.

Works with Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, and any MCP-compatible client.

---

⚡ Quick Start

Claude Code

claude mcp add media-mcp -- uvx media-mcp-server

Claude Desktop / Cursor / VS Code

Add to your MCP config: ``json { "mcpServers": { "media-mcp": { "command": "uvx", "args": ["media-mcp-server"] } } } ``

That's it. No API keys, no accounts, no environment variables.

---

🛠️ Tools

Image Tools

| Tool | Description | |------|-------------| | get_image_info | Get dimensions, format, color mode, file size, EXIF data | | resize_image | Resize by dimensions or scale factor with aspect ratio control | | convert_image | Convert between PNG, JPEG, WebP, GIF, BMP, TIFF, ICO, AVIF | | compress_image | Optimize file size with quality and max dimension controls | | crop_image | Crop to specific pixel coordinates | | create_thumbnail | Generate thumbnails with size control | | strip_metadata | Remove all EXIF/metadata for privacy | | rotate_image | Rotate by any angle with optional expansion | | flip_image | Mirror horizontally or vertically | | apply_filter | Apply blur, sharpen, grayscale, emboss, contour, and more |

Video Tools (requires ffmpeg)

| Tool | Description | |------|-------------| | get_video_info | Get duration, resolution, codec, bitrate, FPS, audio info | | extract_frames | Pull frames at regular intervals | | convert_video | Convert between MP4, WebM, MOV, AVI, GIF, MKV |

---

💬 Example Usage

Once connected, just ask your AI:

"Resize screenshot.png to 800px wide"

"Convert all the PNGs in this folder to WebP"

"Strip the EXIF data from photo.jpg for privacy"

"Compress this image to under 500KB"

"Extract a frame every 5 seconds from demo.mp4"

"What are the dimensions of banner.png?"

"Make a grayscale version of logo.png"

"Create a 128x128 thumbnail of product-photo.jpg"

---

📦 Installation

Using uvx (recommended — zero install)

uvx media-mcp-server

Using pip

pip install media-mcp-server

With video support

pip install media-mcp-server[video]

Note: Video tools require ffmpeg to be installed on your system. Install it from ffmpeg.org or via your package manager: ``bash # macOS brew install ffmpeg # Ubuntu/Debian sudo apt install ffmpeg # Windows (with Chocolatey) choco install ffmpeg ``

---

🔧 Configuration

Config file locations

<details> <summary><b>Claude Desktop</b></summary>

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "media-mcp": {
      "command": "uvx",
      "args": ["media-mcp-server"]
    }
  }
}

</details>

<details> <summary><b>Claude Code</b></summary> ``bash claude mcp add media-mcp -- uvx media-mcp-server `` </details>

<details> <summary><b>Cursor</b></summary>

Settings → MCP Servers → Add: ``json { "media-mcp": { "command": "uvx", "args": ["media-mcp-server"] } } `` </details>

<details> <summary><b>VS Code</b></summary>

Add to .vscode/mcp.json: ``json { "servers": { "media-mcp": { "command": "uvx", "args": ["media-mcp-server"] } } } `` </details>

---

🧪 Development

git clone https://github.com/Adityaaery20/media-mcp.git
cd media-mcp
pip install -e ".[dev]"
pytest

Test with MCP Inspector

npx @modelcontextprotocol/inspector uvx media-mcp-server

---

📋 Supported Formats

Images: PNG, JPEG, WebP, GIF, BMP, TIFF, ICO, AVIF

Videos: MP4, WebM, MOV, AVI, GIF, MKV (requires ffmpeg)

---

🗺️ Roadmap

• [ ] Batch operations (process entire directories)
• [ ] Image watermarking
• [ ] PDF to image conversion
• [ ] OCR (text extraction from images)
• [ ] Audio extraction from video
• [ ] Image collage/montage creation
• [ ] Smart crop (content-aware)
• [ ] SVG rasterization

---

📄 License

MIT — do whatever you want with it.

---

🤝 Contributing

Contributions welcome! Please open an issue first to discuss what you'd like to add.

---

Built with FastMCP and Pillow