home / mcp / memvid mcp server

Memvid MCP Server

Exposes Memvid video memory features to MCP clients for text/pdf encoding, semantic search, and chat over multiple connections.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "angrysky56-memvid_mcp_server": {
      "command": "uv",
      "args": [
        "run",
        "python",
        "memvid_mcp_server/main.py"
      ],
      "env": {
        "PYTHONPATH": "/home/ty/Repositories/memvid_mcp_server",
        "PYTHONWARNINGS": "ignore"
      }
    }
  }
}

You deploy Memvid MCP Server to expose video memory encoding and semantic search capabilities to MCP clients. It lets you encode text, PDFs, and other content into a searchable video memory format, enabling natural-language queries and chat interactions with your encoded data.

How to use

You interact with the Memvid MCP Server through an MCP client. Start the local MCP process, then add content, build a video memory, and use search or chat features to retrieve and discuss your encoded data.

How to install

Prerequisites: Python 3.10 or higher and the uv package manager.

cd /memvid_mcp_server
uv venv --python 3.12 --seed
source .venv/bin/activate
```

```bash
uv add -e .

Configuration and operation notes

The server runs locally via a standard MCP client configuration. Use the following local, stdio-based command to launch the server from your MCP client setup.

{
  "mcpServers": {
    "memvid_mcp": {
      "type": "stdio",
      "name": "memvid_mcp",
      "command": "uv",
      "args": [
        "run",
        "python",
        "memvid_mcp_server/main.py"
      ],
      "env": {
        "PYTHONPATH": "/home/user/memvid_mcp_server",
        "PYTHONWARNINGS": "ignore"
      }
    }
  }
}

Additional setup: Claude Desktop integration

If you use Claude Desktop as your MCP client, you can load the server configuration so Claude Desktop can start and manage the server.

{
  "mcpServers": {
    "memvid-mcp-server": {
      "command": "uv",
      "args": [
        "run",
        "python",
        "memvid_mcp_server/main.py"
      ],
      "env": {
        "PYTHONPATH": "/home/you/memvid_mcp_server",
        "PYTHONWARNINGS": "ignore"
      }
    }
  }
}

Testing the server (optional)

Run the server to verify it starts correctly and responds to MCP client requests.

uv run python memvid_mcp_server/main.py

Security and best practices

Maintain proper file-system permissions for added content and ensure your Docker-based H.265 encoding (if used) runs in a controlled environment. Enable debug logging only during troubleshooting by setting the appropriate environment variable as needed.

Troubleshooting and logs

If you encounter issues, check the logs for errors and verify the server state with the get_server_status tool from your MCP client.

Notes on logging and reliability

The server redirects standard output to avoid interfering with the MCP JSON-RPC protocol. Diagnostic details are sent to standard error, while structured responses are returned to Claude Desktop or other MCP clients.

Available tools

get_server_status

Check current status and version information of the Memvid MCP server.

add_chunks

Add a list of text chunks to the encoder for later video memory construction.

add_text

Add a single text document to the encoder with optional metadata.

add_pdf

Process and add content from a PDF file to the encoder.

build_video

Create the video memory representation from all added content, with options for codec and progress visibility.

search_memory

Perform semantic search against the built video memory and return top results.

chat_with_memvid

Engage in a conversation with the encoded knowledge base.