Trending on GitHub with 19.5K+ Stars

MEMPALACE

The definitive guide to AI memory systems and the ancient art of Memory Palace. From 170-token recall to 2,500-year-old technique.

0
GitHub Stars
0
Recall Rate
0
Token Startup
0
Cost to Run
Choose Your Path

Two Worlds, One Palace

Whether you're building AI agents or training your own memory, the palace awaits.

Architecture

The Palace Structure

Memories organized like rooms in a palace. Navigate from wing to drawer in milliseconds.

Wings

Top-level categories that organize memories by domain or project.

Like wings of a palace housing different knowledge areas

Halls

Thematic groups within a wing, each containing related memories.

Spacious halls where related conversations gather

Rooms

Individual conversation contexts with focused memory scope.

Private rooms for specific topics and threads

Closets

Clustered memory storage for efficient retrieval and organization.

Organized storage keeping memories accessible

Drawers

Individual memory entries with metadata and recall scoring.

The atomic unit: one memory, perfectly indexed

Recall Engine

170-token startup, hierarchical search, optional Haiku rerank.

96.6% recall rate without any API calls

Compare

AI Memory Framework Comparison

How MemPalace stacks up against other AI memory solutions.

MemPalaceBEST
Recall Rate
96.6%
Startup Cost
170 tokens
Pricing
Free
Runs Locally
Yes
License
MIT
Mem0
Recall Rate
~85%
Startup Cost
~2K tokens
Pricing
API costs
Runs Locally
Partial
License
Apache 2.0
Zep
Recall Rate
~80%
Startup Cost
~5K tokens
Pricing
SaaS pricing
Runs Locally
No
License
Proprietary
LangChain Memory
Recall Rate
~70%
Startup Cost
Variable
Pricing
API costs
Runs Locally
Partial
License
MIT
FAQ

Frequently Asked Questions

MemPalace is an open-source AI agent long-term memory system created by Milla Jovovich. It organizes AI conversation memories into a spatial hierarchy (Wings → Halls → Rooms → Closets → Drawers), achieving 96.6% recall with only 170 tokens at startup. It runs entirely locally with zero API costs.

MemPalace achieves higher recall (96.6% vs ~80-85%) with lower startup cost (170 tokens vs 2K-5K). It runs fully locally with no API dependency, while Mem0 and Zep require external services. MemPalace uses a spatial metaphor for organization, making memory retrieval more intuitive.

The Memory Palace, or Method of Loci, is a 2,500-year-old mnemonic technique where you mentally place items to remember in specific locations within an imagined building. By mentally walking through the building, you can recall items in order. It's used by memory champions worldwide.

Yes. MemPalace is MIT-licensed and designed to integrate with any AI agent framework. Install via pip (pip install mempalace), initialize with a few lines of code, and your agent gets persistent long-term memory across conversations.

No. mempalace.info is an independent information hub providing comprehensive guides, comparisons, and tutorials about the MemPalace AI system and the Memory Palace technique. We are not affiliated with the official MemPalace project.

Choose a familiar place (your home works best), identify 5-10 distinct locations along a path, then place vivid mental images at each location representing what you want to remember. Walk the path mentally to recall. Start small and expand as you get comfortable.

Stay Updated

Enter the Palace

Get the latest on AI memory systems, Memory Palace techniques, and MemPalace updates. No spam, unsubscribe anytime.