ChatGPT Memory MCP Server

What is ChatGPT Memory MCP Server?

ChatGPT Memory MCP Server is a tool that exposes ChatGPT's memory as a data source for other language models via the Model Context Protocol (MCP), enabling shared memory outside a single model provider. It is released under the MIT License.

How to use ChatGPT Memory MCP Server?

Clone the repository, install dependencies with pip install -r requirements.txt, set environment variables (Chrome executable path or None), then link the server with an MCP client (e.g., Claude Desktop) by editing the client’s configuration with the appropriate command and arguments.

Key features of ChatGPT Memory MCP Server

• Shares ChatGPT’s memory across different AI models
• Provides access to memory via Model Context Protocol
• Uses Chrome to interact with ChatGPT’s memory
• Open-source under MIT License

Use cases of ChatGPT Memory MCP Server

• Enabling another LLM (like Claude) to read from ChatGPT’s memory
• Creating cross-model workflows that leverage ChatGPT’s stored context
• Building applications that require a persistent shared memory across providers

FAQ from ChatGPT Memory MCP Server

How does this server access ChatGPT’s memory?

It uses a Chrome browser instance to interact with ChatGPT’s memory system, requiring the path to your Chrome executable (or setting it to None for default).

Which applications can I use this MCP server with?

The README provides an example configuration for the Claude Desktop App; any MCP‑compatible client should work.

What runtime dependencies are required?

Python packages listed in requirements.txt and a Chrome browser installation are needed.

Does this server store or modify ChatGPT’s memory directly?

The README only describes reading memory data via MCP; it does not explicitly state whether writing or permanent storage is supported.

Is authentication or API keys required?

No API keys are mentioned; the server relies on a local Chrome instance accessing ChatGPT.