MemGPT
Implements a multi-provider memory system for LLMs, enabling conversation history management and seamless model switching across O
This TypeScript-based MCP server implements a memory system for LLMs, enabling conversation history management across multiple providers. Developed as an open-source project, it offers tools for chatting, retrieving and clearing memory, and switching between LLM providers and models. The implementation supports OpenAI, Anthropic, OpenRouter, and Ollama, with a focus on the latest Claude models. It uses SQLite for persistent storage and includes features like unlimited memory retrieval. This server is particularly useful for building AI assistants that require long-term memory, multi-provider flexibility, and seamless model switching, making it suitable for applications in customer support, content creation, and interactive AI experiences.
Source
Repository: https://github.com/vic563/memgpt-mcp-server
Maintain MemGPT?
Let people know it's listed here — add the badge (live metrics, light/dark aware) or a plain link to your README or docs.
[MemGPT on getagentictools](https://getagentictools.com/mcp/vic563-memgpt-mcp-server?ref=badge) Related mcp
Local by Flywheel Automatically detects and connects to Local by Flywheel WordPress development environments by scanning running processes to find a
CodeRAG Parses TypeScript, JavaScript, Java, and Python codebases into a Neo4J graph database to extract code entities, relationships, and
Lenses Explore, transform, and join data in Kafka topics across multiple clusters
PostgreSQL Advanced Provides full read-write PostgreSQL database access with transaction management, connection pooling, and safety features that requ
RSS Feed Manager Enables AI to access and manage RSS feed content by parsing OPML files, fetching articles, and filtering content from subscribed s
YDB Provides a bridge between AI and YDB databases, enabling natural language interactions for executing SQL queries, exploring schema