
Persistent multimodal context and memory for LLM agents via MCP.
Visit Mcp Context ServerMcp Context Server is a high-performance open-source Model Context Protocol (MCP) server designed for AI agents and LLM applications. It provides persistent multimodal context storage for text and images, enabling advanced memory features such as thread-based context sharing, automatic LLM-powered summarization, semantic/vector/hybrid search, metadata filtering, and batch operations. Integrating with SQLite or PostgreSQL, and supporting embedding and summary models via Ollama, OpenAI, and others, it is production-ready and can be used with Claude, Cursor, LangGraph, and any MCP-compatible client. Ideal for developers building sophisticated, context-aware, multi-agent AI workflows.
Visit Mcp Context Server's official website for product details and getting started.