AI Agents
motorhead
A memory and retrieval server built for LLMs.
★ 917 starsApache-2.0
View guide →// agents · MemTensor/MemOS
MemOS
MemOS is an open-source ai agents project by MemTensor (MemTensor/MemOS on GitHub). It has 11,763 GitHub stars and 1,069 forks, is released under the Apache-2.0 license, and was last updated on GitHub on 8 October 2026.
Self-evolving memory OS for LLMs and agents: persistent graph memory, hybrid retrieval, skill reuse.
★ 11.8k stars⑂ 1.1k forksApache-2.0updated 2026-10
▦ visual — this project's README ships no embeddable screenshot,
so here's a placeholder. The guide below is complete regardless.
so here's a placeholder. The guide below is complete regardless.
// project facts
Project facts
- Repository
- MemTensor/MemOS
- GitHub stars
- 11,763
- Forks
- 1,069
- License
- Apache-2.0
- Last updated on GitHub
- 8 October 2026
- Category
- AI Agents
- Listing
- Full install guide
- Official site
- Official site
// how it works
How it works
- Pick your deployment shape
Choose the hosted Cloud API for zero infrastructure, self-hosting for your own servers, or the local plugin for 100% on-device memory in OpenClaw, Hermes or DeepSeek Harness.
- Set it up
For self-hosting, clone the repo, copy docker/.env.example to .env and fill in your LLM and embedder API keys; for cloud, create an API key in the dashboard.
- Start the service
Bring up MemOS with Docker Compose (API, Neo4j and Qdrant) and confirm it answers on port 8000.
- Add and search your first memories
Create a memory cube, add conversation messages through the API, then search them back and watch recall improve future agent turns.
// features
Why people love it
Unified memory API
One API to add, retrieve, edit and delete memory, structured as an inspectable graph rather than an opaque embedding store.
Multimodal and tool memory
Natively stores text, images, tool traces and personas, retrieved and reasoned over together in one system.
Memory cubes
Manage multiple knowledge bases as composable cubes with isolation and controlled sharing across users, projects and agents.
Asynchronous ingestion
The MemScheduler runs memory operations asynchronously, keeping latency low under high concurrency for production stability.
Feedback and correction
Refine memory with natural-language feedback - correcting, supplementing or replacing memories over time, which underpins its self-evolving behavior.
Plugins for agent frameworks
Official cloud and local plugins for OpenClaw, Hermes Agent and DeepSeek Harness add automatic recall before tasks and capture after them, with a local Memory Viewer.
Advertisement
// run it
Run it yourself
- Clone the repository
Clone MemOS for self-hosting; cloud and plugin users can skip the clone.
git clone https://github.com/MemTensor/MemOS.git cd MemOS - Configure your keys
Copy the example environment file and fill in your LLM provider and embedder API keys.
cp docker/.env.example .env # fill in your API keys in .env - Start the service
Launch MemOS with Docker Compose - it starts the API together with Neo4j and Qdrant - then call it on port 8000.
cd docker docker compose up
// use cases
What you can build
- Giving a customer-support agent recall of past tickets and each user's history.
- Building a personal assistant that keeps consistent preferences across months of conversations.
- Letting several collaborating agents share a project memory cube while keeping user memories isolated.
- Adding persistent, self-evolving memory to an OpenClaw or Hermes agent with the local plugin.
Advertisement
// faq
Questions, answered
Who maintains MemOS?
MemOS is developed by MemTensor and contributors; the research is described in the paper 'MemOS: A Memory OS for AI System' (arXiv 2507.03724).
Is MemOS free?
The open-source code is Apache-2.0 licensed and self-hosting is free. The hosted MemOS Cloud API is a managed service with its own key and terms.
Does it need a paid API key?
Self-hosting needs your own LLM and embedder keys (set in .env) for memory extraction and search; the cloud API uses a MemOS API key instead. The local plugin keeps data on-device.
How does MemOS compare with Mem0 or EverOS?
All three give agents long-term memory; MemOS centers on a graph-structured memory OS with composable memory cubes and an async scheduler, and ships first-party plugins for OpenClaw, Hermes and DeepSeek Harness.
How do I get started?
For the fastest path, get a cloud API key from the MemOS dashboard and call the add/search endpoints; for self-hosting, clone, fill .env, and docker compose up.
// keep exploring
Similar tools
AI Agents
evolver
A GEP-powered self-evolving engine for AI agents — auditable evolution with Genes, Capsules and Events.
★ 9.1k starsGPL-3.0
View guide →AI Agents
memory-lancedb-pro
Memory-LanceDB-Pro gives OpenClaw agents hybrid search, smart extraction, and intelligent forgetting
★ 4.5k starsMIT
View guide →AI Agents
Memori
Agent-native memory infrastructure - structured, persistent memory for production agents
★ 17.1k starsNOASSERTION
View guide →AI Agents
LongMemory
LongMemory: local-first memory for AI agents with temporal truth, provenance, and hybrid recall
★ 4.5k starsApache-2.0
View guide →AI Agents
m_flow
M-flow is a bio-inspired cognitive memory engine for AI agents — a graph alternative to flat RAG
★ 4.5k starsApache-2.0
View guide →// alternatives
Alternatives in AI Agents
The most-starred other AI Agents projects in the LibreAI catalogue — the realistic shortlist if MemOS is not the right fit.
- openclaw★ 391.6k
A personal AI that really does things — on any OS, with your data staying yours.
- superpowers★ 294.9k
An agentic skills framework & software development methodology that works.
- ECC★ 275.2k
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
- hermes-agent★ 252.0k
The agent that grows with you — a Python framework for adaptive AI agents.
- deepseek-harness★ 245.5k
DeepSeek's plugin-based agent harness where everything is a plugin.
// more in agents
More in AI Agents
Nearby AI Agents projects in the catalogue, closest in stars to MemOS — keep exploring the shelf either side of it.
- bisheng★ 12.0k
Enterprise LLM app platform: GenAI workflows, RAG, agents, model management, evaluation, fine-tuning.
- phoenix★ 11.7k
Observability and evaluation for AI apps — see what your agents do.
- hexstrike-ai★ 12.5k
MCP server letting AI agents run 150+ cybersecurity tools for authorized pentesting and research.
- pest★ 11.7k
The elegant testing framework for PHP developers and AI agents.
- ida-pro-mcp★ 12.5k
AI-powered reverse engineering assistant that bridges IDA Pro with language models through MCP.
- Loop Engineering★ 11.6k
Loop Engineering — practical patterns, starters and a CLI for designing the loops that prompt and orchestrate AI coding agents, with a Loop Ready score
Get MemOS
Free and open — grab the source, star the repo, and start creating.