Milla Jovovich Built MemPalace — the AI Memory System That Broke Every Benchmark.

MemPalace was created by Milla Jovovich (the actress known for the Resident Evil franchise) and developer Ben Sigman. They built it over several months using Claude Code after Milla Jovovich became frustrated that existing AI memory systems would decide what to remember for her, discarding the reasoning and context she actually needed. The project is now available on Milla Jovovich's GitHub (milla-jovovich/mempalace).

Milla Jovovich co-developed MemPalace using Claude Code, an AI-assisted coding tool. She described the project as a collaboration between her vision for how Milla Jovovich AI memory should work and Ben Sigman's engineering expertise. As she said, "I immediately realized that as an artist who loves to write, Claude could turn my words and ideas into reality." According to Ben Sigman: "By day, she's filming action movies, walking Miu Miu fashion shows, and being a mom. By night she's coding." This Milla Jovovich AI vision is what sets MemPalace apart from other solutions.

MemPalace (sometimes searched as MemPlace or Milla Jovovich MemPlace) is an open-source AI memory system that gives large language models persistent, cross-session memory. It stores all conversation data verbatim and uses vector search to retrieve relevant memories. It runs entirely locally with zero API costs, using ChromaDB and SQLite. The system draws inspiration from the ancient "method of loci" mnemonic technique. You can find it on GitHub under GitHub MemPalace or MemPalace GitHub.

MemPalace claims 100% on LongMemEval in hybrid mode (with optional Haiku reranking) and 96.6% in raw mode (zero API). However, these scores have been questioned by independent reviewers. The 100% hybrid score was achieved after targeted fixes for specific failing questions, and some methodology choices (like top_k settings) have been debated. The raw 96.6% LongMemEval score is generally considered more credible.

Yes. MemPalace is MIT licensed and runs entirely on your local machine. The raw search mode requires zero API keys and zero cost. The optional hybrid mode uses a cloud LLM for reranking at approximately $0.001 per query. You can download it from MemPalace GitHub or install via PyPI.

The key difference is philosophical: MemPalace stores everything verbatim and uses vector search, while Mem0 uses an LLM to extract key facts and discards the rest. MemPalace scores higher on LongMemEval benchmarks (96.6-100% vs ~85%) and costs nothing, but Mem0 has $24M in funding, enterprise support, and a more mature ecosystem.

Install MemPalace with pip install mempalace, then run mempalace init. It requires Python 3.9+. For Claude Code integration, add the MCP server config to your settings. The latest secure version is 3.0.0 according to PyPI. You can also clone directly from GitHub MemPalace at milla-jovovich/mempalace.

AAAK is a custom lossless compression dialect that MemPalace uses to compress stored memories approximately 30x. Any LLM can read AAAK natively without a decoder. However, independent testing suggests that using AAAK may reduce retrieval accuracy from 96.6% to approximately 84.2%, which is a tradeoff worth considering. The AAAK compression system is one of MemPalace's unique innovations.