M3 Memory

Persistent, local memory for MCP agents.

Your agent forgets everything between sessions. M3 Memory fixes that. Install it, add one line to your MCP config, and your agent remembers across sessions, detects contradictions, and keeps its own knowledge current — all on your hardware, fully offline.

Works with Claude Code, Gemini CLI, Aider, and any MCP-compatible agent.

Install

pip install m3-memory

Add to your MCP config:

{
  "mcpServers": {
    "memory": { "command": "mcp-memory" }
  }
}

Requires a local embedding model. Ollama is the easiest:

ollama pull nomic-embed-text && ollama serve

Prefer a GUI? LM Studio works too — load any embedding model (e.g. nomic-embed-text-v1.5) and start its server (defaults to port 1234).

Restart your agent. Done.

What happens next

You're at a coffee shop on your MacBook, asking Claude to debug a deployment issue. It remembers the architecture decisions you made last week, the server configs you stored yesterday, and the troubleshooting steps that worked last time — all from local SQLite, no internet required.

Later, you're at your Windows desktop at home with Gemini CLI, and it picks up exactly where you left off. Same memories, same context, same knowledge graph. You didn't copy files, didn't export anything, didn't push to someone else's cloud. Your PostgreSQL sync handled everything in the background the moment your laptop hit the local network.

Why this exists

Most AI agents don't persist state between sessions. You re-paste context, re-explain architecture, re-correct mistakes. When facts change, the agent has no mechanism to update what it "knows."

M3 Memory gives agents a structured, persistent memory layer that handles this.

What it does

Persistent memory — facts, decisions, preferences survive across sessions. Stored in local SQLite.

Hybrid retrieval — FTS5 keyword matching + semantic vector similarity + MMR diversity re-ranking. Scored and explainable.

Contradiction handling — conflicting facts are automatically superseded. Bitemporal versioning preserves the full history.

Knowledge graph — related memories linked automatically on write. Eight relationship types, 3-hop traversal.

Local and private — embeddings generated locally. No cloud calls. No API costs. Works offline.

Cross-device sync — optional bi-directional delta sync across SQLite, PostgreSQL, and ChromaDB. Same memory on every machine.

Who this is for

... View full README on GitHub