Introduction
llm-mem is a memory server for AI agents. It gives your AI assistant persistent, searchable memory that works across sessions — conversations, codebases, documents, and anything else you want it to remember.
It’s a single self-contained binary written in Rust. No databases to install, no cloud services required, no API keys needed by default. Everything runs locally on your machine, and data never leaves your computer.
Warning
Beta software —
This is beta software. Expect rough edges and breaking changes. Use it to try things out, not for anything important.
How it works
When connected to an AI assistant via the Model Context Protocol (MCP), llm-mem exposes a set of tools the assistant can call to:
- Store raw text, documents, or AI-extracted facts as memories
- Search by meaning (semantic search) or by keyword
- Browse chronologically with timeline tools
- Navigate a layered pyramid from raw content up to abstract insights
- Manage isolated memory banks, backups, and document uploads
The assistant calls these tools just like any other MCP tool — no special prompting required.
What makes it different
| Feature | llm-mem |
|---|---|
| Local-first | Runs entirely on your machine by default. No cloud, no accounts, no data leaves your box. |
| Embedded vector store | Uses LanceDB embedded. No separate database server. |
| Layered memory | L0 raw → L1 summaries → L2 connections → L3 concepts → L4 insights, with auto-abstraction. |
| AI in-process | Bundled llama.cpp for local LLM, fastembed for local embeddings. Zero-config. |
| MCP-native | Speaks MCP stdio. Drop into opencode, VS Code Copilot, or any MCP client. |
| Standalone CLI | Includes an interactive REPL and single-command mode for shell scripting. |
Next steps
- New here? Start with What is llm-mem? for the conceptual overview.
- Ready to install? Jump to Installation.
- Want to skim the tool surface? See Tools overview.