Memoria 1.0: a local, model-agnostic LLM memory layer hitting 89.8% Recall@1 on LongMemEval
kitkatz69 · reddit · 2026-10-06
A developer released Memoria 1.0.0, a local-first, LLM-agnostic memory layer for LLM apps that runs fully offline with no API keys.
- Runs on tiny hardware: benchmarked on a GPU-less Intel Celeron N4020 (1.10 GHz, 3.7 GiB RAM); peak RSS for the full LongMemEval workload was just 2.65 GiB (580 MiB per query).
- Benchmarks (LongMemEval-S, 468/470 evaluable questions): Recall@1 89.8%, Recall@5 97.9%, Recall@10 98.9%, Recall@50 99.6%; Session NDCG@10 0.9257.
- Retrieval architecture: not a single vector DB — FAISS, BM25, graph retrieval, phrase matching, attribute retrieval and temporal retrieval run in parallel, fused via multi-signal ranking; temporal retrieval is independently implemented for ablation.
- Also ships: GitHub repo and Obsidian vault ingestion, MCP support, CLI/TUI/GUI/API, persistent local storage, LongMemEval/LoCoMo benchmark tooling, and a plugin system with 11 subsystems and 34 hooks plus an interactive plugin generator.
- Install via pip install kitzkatz-memoria; source and docs are on GitHub, with feedback sought from local-agent builders.
More from coding & agent
- Creator hands repetitive workflow to Codex and GPT-6 Astra, keeps creative calls — socialwithaayan · 2026-10-06
- Monica CRM MCP Server wraps REST API with 21 natural-language tools — modelcontextprotocol · 2026-10-06
- React Compiler core member's go-to prompt: make AI restate your goals before it works — sujingshen · 2026-10-06
- dotey explains Codex Project vs Claude Projects: forum boards vs Slack channels — dotey · 2026-10-06
- A Planted 'P.S.' Fooled Jev, TypeSafe's New Decision Model — a Simple Rule Caught It — Internal-Lie-5197 · 2026-10-06
- Claude Code's New Dreaded Message: 'Compacted While Idle, Before the Prompt Cache Expired' — dSebastien · 2026-10-06