7.0.2 Mnemosyne OS now ships a 20-tool MCP server and a core with zero third-party dependencies. Read the changelog
Mnemosyne OS 7.0.2 Zero-cloud memory layer

Zero-cloud memory for AI agents.

A local-first memory engine that runs on the Python standard library alone. Sub-millisecond writes, a hash-chained audit ledger, and pure SQLite storage — no vector cloud, no vendor lock-in, no data leaving your machine.

Mounts into Claude Code/ Cursor/ OpenWebUI/ Hermes Agent/ DeepSeek Harness/ LangChain/ Ollama see all →
Product

Built for developers who want proof, not promises

Mnemosyne gives agents persistent memory without a pipeline change. Less redundant context, lower token cost, measurably faster responses.

Try it now
How it works

Retain anything. Mnemosyne learns.

Three stages run on every write and every read. No configuration, no boilerplate, no pipeline to rewire.

Architecture

BEAM — bi-layer episodic and associative memory.

Things that happened in time and things that relate to each other are structurally different. BEAM stores them differently, then assembles a single injection under a token budget.

WORKING EPISODIC TRIPLESTORE CONTEXT RECALL GRAPH fusion t = live
write path recall path temporal chain

Pure wireframe — every line and packet is drawn in-browser with SVG. No library, no canvas, no colour noise.

Use cases

Memory that adapts to your domain

Mnemosyne helps AI remember what matters.

Integrations

Anything that speaks MCP already works.

Mnemosyne exposes its 20 tools over the standard Model Context Protocol, so it mounts without a bespoke adapter. Every transport below has a runnable or a config in the repository.

Every entry above rides the public MCP surface — stdio or Streamable HTTP. No forked client, no vendor plugin, no private SDK.

New algorithm

Benchmarking Mnemosyne

Single-pass hierarchical distillation. Multi-signal retrieval. Benchmarked across LoCoMo, LongMemEval and BEAM — the three benchmarks the memory field is actually judged on.

Mnemosyne OS 7.0.2 Mem0

Source note — Mnemosyne figures come from this project's 7.0.2 benchmark and retrieval-quality reports, measured on a real local corpus. Competitor figures are order-of-magnitude estimates drawn from their public documentation and pricing pages; they are not same-hardware head-to-head results. The chart shows architectural order of magnitude, not a lab comparison.

Enterprise

Built for enterprise. Designed for control.

Memory at scale is infrastructure. Mnemosyne gives teams governance, portability and full observability so engineers spend time building, not recovering lost context.

Compliance is documented rather than asserted: a mapping for HIPAA, 等保, GDPR and PIPL ships with the repository, alongside a SHA-256 hash-chain ledger that makes tampering detectable rather than deniable.

Pricing

The open-source tier is free forever, and not crippled.

The engine is MIT. You pay for multi-tenancy, cloud coordination and compliance delivery — never to unlock a feature.

Final pricing subject to commercial confirmation. A complete free-tier cloud deployment package (managed free VM + free embedding API + free HTTPS tunnel) ships inside the repository.

Resources

Latest from the repository

Give your agent a memory it actually owns.

No account, no credit card, no network. One pip command and your machine has a memory engine that will not forget you.

Read the source

python 3.8+ · windows / macOS / linux · offline capable · no telemetry