memory-os vs MemOS
A 7-layer memory operating system for Hermes Agent: Qdrant vectors, structured facts, fabric recall, an auto-curated wiki, and surgical context injection. — versus — MemTensor's memory OS for LLM agents: one API over graph-structured, multi-modal memory with hybrid retrieval and skill evolution — hosted, self-hosted (Neo4j + Qdrant) or local plugins.
Both pitch a layered 'memory OS' with a Hermes Agent integration — memory-os is a 7-layer Qdrant stack built for Hermes, MemOS a multi-harness core with local SQLite plugins.
| memory-os | MemOS | |
|---|---|---|
| Stars | 1.4k | 12k |
| Forks | 128 | 1.1k |
| Language | Python | TypeScript |
| License | MIT | Apache-2.0 |
| Last activity | 8 days ago | 5 days ago |
| Topics | memory | memory, skills |
| Curated connections | 6 | 5 |
memory-os — the curator's take
The most architecturally ambitious take on agent memory we've mapped: seven distinct layers from raw vectors to an auto-curated wiki, each with its own recall path, so the agent gets the RIGHT kind of memory injected rather than a similarity dump. The catch is coupling: it's built FOR Hermes Agent — adopting the architecture elsewhere means porting, not installing; and 7 layers is real operational surface for a ~1.3k-star project.
MemOS — the curator's take
Pick MemOS when you want a full memory stack rather than a vector wrapper: graph-structured memories you can inspect and correct in natural language, isolated or shared 'memory cubes' across users and agents, async ingestion, and traces that crystallize into reusable skills. The fastest path is a plugin — local SQLite for Hermes, OpenClaw or DeepSeek Harness, zero infra. Self-hosting the service means running Neo4j and Qdrant plus LLM and embedder keys, which is heavy for a single-user bot. Read the headline LoCoMo/LongMemEval numbers with care: the comparison runs on OmniMemEval, MemTensor's own harness. Skip it if you only need session recall for one coding agent — claude-mem or engrim is far less machinery.