# Hermes Agent vs MemOS

> Side-by-side comparison of two projects in the OpenClaw ecosystem.

Canonical HTML: https://shelldex.com/compare/hermes-vs-memos/
JSON: https://shelldex.com/api/compare/hermes-vs-memos.json

| Field | Hermes Agent | MemOS |
| --- | --- | --- |
| Tagline | The agent that grows with you | AI memory OS for agent systems |
| Language | python | python |
| Status | active | active |
| Stars | 163k | 9.3k |
| Forks | 26728 | 845 |
| Open issues | 13267 | 173 |
| License | MIT | Apache-2.0 |
| Last commit | 2026-05-23T03:15:11Z | 2026-05-22T13:36:57Z |
| 7d growth | 11664 | 235 |
| MCP support | yes | no |
| Requires LLM | yes | yes |
| Platforms | server, messaging | server |
| Categories | memory, self-hosted, multi-platform | memory, infrastructure |
| Health score | 75/100 | 95/100 |
| Leaderboard rank | #2 | #15 |

## Key differences

- Hermes Agent has 17x more stars (163k vs 9.3k), indicating significantly broader adoption.
- Hermes Agent is growing faster with +12k stars this week vs +235 for MemOS.
- Hermes Agent uses the MIT license while MemOS uses Apache-2.0.
- Hermes Agent has MCP (Model Context Protocol) support while MemOS does not.
- Hermes Agent focuses on self-hosted, multi-platform while MemOS targets infrastructure.
- MemOS scores higher on project health (maintenance activity, issue management, release cadence).

## Descriptions

### Hermes Agent

Persistent personal AI agent by Nous Research with multi-level growing memory, self-authored skills, and multi-platform messaging. Learns your projects and builds reusable knowledge over time.

### MemOS

Persistent skill memory for cross-task reuse and evolution — built for OpenClaw and similar agent frameworks