MnemOS
A universal, persistent memory layer for AI assistants — designed, built, and open-sourced solo.
AI assistants forget everything between sessions. MnemOS solves that by extracting facts from conversations via LLM, embedding them with bge-m3, and injecting relevant context before each new message — invisibly, across ChatGPT and Claude. Memories are stored per-user rather than per-platform, so context follows the person across tools instead of resetting every session.
01 / Onboarding
This is the actual setup wizard, rebuilt as a working component with synthetic data — click through it below. Check the server, pick an engine, download a model, and finish setup.
Universal Memory for AI — set up in 2 minutes
Connect to MnemOS Server
MnemOS needs a local server running on your machine to store and retrieve memories.
Checking server...
Looking for MnemOS at localhost:8765
02 / Extension popup
The Chrome extension popup — remove a memory, switch the extraction model, and apply it.
Memories 5
User wants to learn Rust for inference coding
User wants to learn paragliding
User prefers VS Code for development
User will install Rust using rustup
User needs to understand ownership and borrowing in Rust
Extraction Model
03 / Dashboard
Overview, Memories, Sessions, and Settings — all four tabs are live. Search and delete memories, switch tabs, download a model.
Total Memories
13
Sessions
7
Semantic
52
Episodic
6
By Application
Recent Memories
User wants to learn paragliding
User wants to learn Rust for inference coding
User needs to understand ownership and borrowing in Rust
User will install Rust using rustup