Home

SuperLocalMemory V3.6

Save up to 90% on every LLM API call. Cache. Compress. Remember. The only local-first memory system that SKIPS repeat calls (100% saved), SHRINKS prompts 60-95%, and REMEMBERS everything — locally, for free.

SuperLocalMemory gives AI assistants persistent memory across sessions. v3.6 "Optimize" adds Cache + Compress + Align — a local-first LLM cost-reduction layer alongside SLM's existing memory engine. One install. One daemon. One UI.

v3.6 "Optimize" — Cache. Compress. Align. ✨ NEW

Skip repeat LLM calls (100% saved), shrink prompts 60-95%, and stabilize prefixes for native provider KV-cache discounts. Local-first. Fail-open. AES-256-GCM encrypted. One command: slm wrap claude. Read more →

v3.4.5 "Scale-Ready" — 1 Million Memories. Zero Slowdown.

Tiered storage auto-classifies every memory as active, warm, cold, or archived. Graph pruning removes redundant connections. Optional acceleration backends (CozoDB, LanceDB) for graph + vector operations. Tested on 1.18 million real graph edges with under 2-second recall. Migration is automatic: pip install -U superlocalmemory && slm restart. View details →

V3.3.6: Zero-Friction Hooks — Install Once, Forget Forever

One npm install and your AI memory is fully automatic:

• Auto-recall at session start — your context is there before you ask
• Auto-observe during coding — decisions and changes captured silently
• Auto-save at session end — full summary with git context
• Zero setup — hooks install themselves, no config needed
• Zero risk — every hook fails silently, never blocks your workflow

V3.1: Active Memory — Memory That Learns

SLM learns from your usage patterns and gets smarter over time — at zero token cost. Every recall generates learning signals. After 20+ signals, the system starts optimizing retrieval for YOUR specific patterns. After 200+, a full ML model trains on your data. No other memory system learns without spending LLM tokens. Read more →

Quick Start

pip install -U superlocalmemory    # Install or upgrade
slm restart                         # Restart daemon (auto-migrates)
slm wrap claude                     # Start saving on LLM costs
slm optimize savings --since 1      # See your savings

That's it. Your AI now remembers you — and saves you money.

Research Papers

Paper	Title	Link
Paper 3 (2026)	The Living Brain — Forgetting, Quantization, 7-Channel Retrieval	arXiv:2604.04514
Paper 2 (2026)	Information-Geometric Foundations for Zero-LLM Agent Memory	arXiv:2603.14588
Paper 1 (2026)	Trust & Behavioral Foundations for Multi-Agent Memory	arXiv:2603.02240

---

Part of Qualixar | Created by Varun Pratap Bhardwaj | AI Reliability Engineering