Open-source memory framework for AI agents that stores learned context as searchable markdown files, priced by token usage.
What it does
MemU is an open-source memory system for AI agents that keeps what agents learn as plain markdown files (MEMORY.md, INDEX.md, SKILL.md) rather than opaque storage. It offers hybrid retrieval across files, full provenance back to source chats, and Memory/Response APIs for building always-on, proactive agents.
How to use: MemU offers three ways to leverage our memory system: our managed platform, our enterprise solution, and our open source solution.
Core features
File-based memory stored as markdown
Hybrid search across memory files
Full provenance to source conversations
Memory API and Response API
Open-source components (server, UI)
Multi-user/agent, multimodal memory
Best for
→Giving AI agents persistent memory
→Building 24/7 proactive agents
→AI companions and coding assistants
→Knowledge management for agents
Toolspool rankingby monthly traffic
Reviews
Big-picture takes: what it's for and whether it delivers. High-engagement YouTube videos — not sponsored.
Tutorials
Step-by-step: exactly how to get things done with it.