m3-memory 2026.5.4.6

Local-first persistent memory for MCP agents — 73 tools, 89% LongMemEval, hybrid search, GDPR, zero cloud

![M3 Memory]

M3 Memory

Persistent, local memory for MCP agents.

"Wait, you remember that?" — Stop re-explaining your project to your AI. Give it a long-term brain that stays 100% on your machine.

🚀 New to M3? Start here with our 5-minute "Human-First" guide.

Works with Claude Code, Gemini CLI, Aider, OpenCode, and any MCP-compatible agent. Quick one-line command to have your agent install chat log sub-system which saves verbatim chat log info, before compaction, with zero lag/latency and 100% retrieval recall. Just tell your AI agent "install m3-memory chat log sub-system" and your agent will automatically install it with all the proper hooks with some minimal customization questions from you (you can accept the default answers).

---

📦 Install

curl -fsSL https://raw.githubusercontent.com/skynetcmd/m3-memory/main/install.sh | bash

Installs on macOS or Linux with the single command above. Use this to install on Windows. Use this link to install manually and this to examine the script and what it does.

Claude Code users can also install as a plugin instead — gets you 15 /m3:* slash commands, a memory-curator subagent, and auto-wired hooks:

/plugin marketplace add skynetcmd/m3-memory
/plugin install m3@skynetcmd

Plugin reference · Claude.ai (web/desktop) connector

---

Add to your MCP config:

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

Requires a local embedding model. Ollama maybe the easiest:

ollama pull qwen3-embedding:0.6b && ollama serve

I personally use LM Studio for it's support for Apple's MLX. But, it's your preference.

Qwen3-Embedding-0.6B (1024-dim, Q8 quantized, ~639 MB) is the model M3 Memory is tuned for. But you can set and use other embedders as you wish. For example, you could use nomic-embed-text (768-dim) which also works (with minimal fidelity loss) — set EMBED_MODEL=nomic-embed-text in your environment.

As mentioned, you can use Ollama or LM Studio — load an embedding model and start its server.

Want auto-classification, summarization, and consolidation? Load a small chat model alongside the embedder (e.g. qwen2.5:0.5b via Ollama, or any 0.5–1B instruct GGUF in LM Studio / llama.cpp). M3 auto-selects it; embedding-only features work without it. See docs/QUICKSTART.md → Optional: load a small chat model.

Restart your agent. Done!

---

🔮 What happens next (benefits of use)

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

Autonomous cognitive loop — optional background worker (m3_cognitive_loop.py) that extracts facts, resolves contradictions, and links entities while you sleep. Turns raw chat logs into a refined knowledge graph without human intervention.

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

Hybrid retrieval — FTS5 keyword matching + semantic vector similarity + MMR diversity re-ranking. Automatic, no tuning required.

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

Knowledge graph — related memories linked automatically on write. Nine relationship types, 3-hop traversal. Entity extraction (entity_search, entity_get) supplements the graph with first-class people / places / things resolution.

Zero-config local install — pip install m3-memory plus one line in your MCP config, or mcp-memory install-m3 for a one-command setup that wires settings.json, hooks, and the chatlog subsystem in one shot. SQLite stores everything locally — no external databases, no cloud calls, no API costs. Works offline.

Cross-device sync — optional, easy-to-add bi-directional delta sync via PostgreSQL or ChromaDB, with manifest-driven multi-DB support for fleet deployments. Set one environment variable and your memories follow you across machines.

---

📚 Learn more

🚀 Getting started	👥 Multi-agent orchestration
✨ Core features	🧩 Multi-agent example
🏗️ System design	⚖️ Compare M3 to alternatives (sovereign substrates table)
🔧 Implementation details	⚙️ Configuration
🤖 Agent rules + all 73 tools	🛡️ Compliance & assurance (FISMA, CMMC, GDPR)
🏠 Homelab patterns	🔍 Myths & facts (verify claims about M3)
🗺️ Roadmap

---

🎯 Who this is for

M3 is a good fit if…

🤖 You use coding agents	Claude Code, Gemini CLI, Aider, OpenCode, or any MCP-compatible agent. Non-MCP clients work too via the built-in HTTP proxy.
👥 You run multiple agents	Coordinating Claude + Gemini + a background worker on a shared local store, with handoffs and per-agent scoping.
🛡️ You need compliance primitives	gdpr_forget / gdpr_export as MCP tools, bitemporal valid-time / transaction-time, audit trail, no telemetry.
💾 You want pure local-first	Single-file SQLite. Works offline. No external database, no cloud calls, no API costs by default.
🌐 You want memory across devices	Optional bi-directional delta sync via PostgreSQL or ChromaDB — your data, your hardware.

M3 is not the right tool if…

	Try instead
You're building LangChain / LangGraph / CrewAI pipelines and want framework-native memory	Mem0, LangChain Memory / LangMem
You want a hosted agent runtime with managed scaling, dashboards, and SLAs	Letta, Mem0 Pro
Pure retrieval-accuracy is your only criterion (M3 is mid-pack at 89.0% LME-S)	agentmemory (96.2%), Hindsight
You only need in-session chat context that's discarded after the conversation	Your agent's built-in conversation buffer; M3 is overkill

---

🛡️ Why trust this

73 MCP tools	Memory, search, GDPR, refresh lifecycle — plus agent registry, handoffs, notifications, tasks, entity graph, fact enrichment, and chat-log capture for multi-agent orchestration
193 end-to-end tests	Covering write, search, contradiction, sync, GDPR, maintenance, and orchestration paths
Explainable retrieval	memory_suggest returns vector, BM25, and MMR scores per result
SQLite core	No external database required. Single-file, portable, inspectable
GDPR compliance	gdpr_forget (Article 17) and gdpr_export (Article 20) as built-in tools — see compliance & assurance for FISMA / CMMC alignment too
Self-maintaining	Automatic decay, dedup, orphan pruning, retention enforcement
Audited security posture	Periodic Bandit + pip-audit + secrets-scan reports published under docs/audits/; CI gates on core-dep CVEs
Apache 2.0 licensed	Free. No SaaS tier, no usage limits, no lock-in

🧭 Maturity, honestly. The core (storage, retrieval, GDPR, MCP tools, sync) is stable and covered by the test suite. The newer enrichment + reflector pipeline matured rapidly through 2026-Q2 and has live-fire experience behind it but is still iterating. Production-ready for personal, homelab, and multi-agent developer workflows today. For regulated workloads, do your own evaluation against your specific use case — and we recommend that against any memory tool, not just M3. See docs/MYTHS_AND_FACTS.md for what we don't claim.

---

📊 Benchmarks

89.0% on LongMemEval-S (445/500 correct) — a 500-question evaluation of long-horizon conversational memory. Without oracle metadata: 74.8% (smart retrieval) to 68.0% (fixed-k baseline).

Question type	n	Accuracy
single-session-user	70	91.4%
single-session-assistant	56	94.6%
single-session-preference	30	93.3%
multi-session	133	85.0%
temporal-reasoning	133	86.5%
knowledge-update	78	92.3%
Overall	500	89.0%

Full methodology, ablations, and honest caveats: benchmarks/longmemeval/LME-S_Benchmarking_Report.md. LoCoMo audit pending — see benchmarks/locomo/README.md.

🔍 Verifying claims about M3. If a third-party AI assistant has described M3 with features or scores that don't match what's documented here, it's almost certainly hallucinating. See docs/MYTHS_AND_FACTS.md for the source-of-truth list of what M3 actually implements (and what it doesn't).

---

🧰 Core tools

Most sessions use three tools. The rest is there when you need it.

Tool	Purpose
memory_write	Store a fact, decision, preference, config, or observation
memory_search	Retrieve relevant memories (hybrid search)
memory_update	Refine existing knowledge
memory_suggest	Search with full score breakdown
memory_get	Fetch a specific memory by ID

All 73 tools are documented in docs/AGENT_INSTRUCTIONS.md and the full inventory lives in docs/MCP_TOOLS.md.

---

🤖 For AI agents

M3 Memory exposes 73 MCP tools for storing, searching, updating, and linking knowledge — including conversation grouping, a refresh lifecycle for aging memories, agent registry, handoffs, notifications, tasks, entity-graph extraction, fact enrichment, and chat-log capture for multi-agent orchestration. Any MCP-compatible agent can use them automatically.

To teach your agent best practices (search before answering, write aggressively, update instead of duplicating), drop the compact rules file into your project:

examples/AGENT_RULES.md

Full tool reference with all parameters and behaviors: docs/AGENT_INSTRUCTIONS.md

---

🪄 Let your agent install it

Already inside Claude Code or Gemini CLI? Paste one of these prompts:

Claude Code:

Install m3-memory for persistent memory. Run: pip install m3-memory
Then add {"mcpServers":{"memory":{"command":"mcp-memory"}}} to my
~/.claude/settings.json under "mcpServers". Make sure Ollama is running
with qwen3-embedding:0.6b. Then use /mcp to verify the memory server loaded.

Gemini CLI:

Install m3-memory for persistent memory. Run: pip install m3-memory
Then add {"mcpServers":{"memory":{"command":"mcp-memory"}}} to my
~/.gemini/settings.json under "mcpServers". Make sure Ollama is running
with qwen3-embedding:0.6b.

After install, test it:

Write a memory: "M3 Memory installed successfully on [today's date]"
Then search for: "M3 install"

Add the chat log subsystem

Want auto-capture of every Claude Code / Gemini CLI / OpenCode / Aider conversation into a searchable, promotable chat log store? Once m3-memory is wired up, just say:

Install the m3-memory chat log subsystem.

The agent runs bin/chatlog_init.py, wires the host-agent hook, and installs the embed sweeper schedule. See docs/CHATLOG.md for the architecture and ops guide.

---

🎬 See it in action

Contradiction detection

Hybrid search with scores

Cross-device, cross-platform sync

---

💬 Community

    Contributing · Good first issues

---