Source checked · Research
MemOS
What the publisher saysSelf-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task skill reuse, with 35.24% token savings.
Potential use: Research teams that need to build and evaluate long-term memory for agents.
Limitations: Confidential documents or project memory until storage, retention, and deletion controls are clear.
- Public use
- 10,220 GitHub stars · 932 forks
- Activity
- Release v2.0.23 on 9 Jul 2026 · repository updated 15 Jul 2026
- Community
- 49 open issues · Discussions available
- Checked
- 15 Jul 2026
Source facts only. Not tested or recommended by Agent Blocks.