Local-first memory system that gives every AI coding agent one shared, searchable brain on your machine.
Homepage hero - rent the LLM, own the memory
1 / 2Overview
What is SuperLocalMemory?
Agent memory with Laya, Jev + Google Gemma
SuperLocalMemory is a local-first memory system that gives every AI agent one shared, governed memory brain. Claude Code, Cursor, VS Code, Windsurf, Codex, ChatGPT and any MCP/CLI agent can remember, recall and search facts, screenshots, photos and PDFs from a single local store. Its answer-checking step grades every recall as supported, related or no answer — and says 'I don't have that' instead of guessing. Judging happens locally on Apple Silicon (Laya) or online with explicit consent (Jev). Governance features include roles, scopes, workspaces, verifiable full deletion, hash-chained audit logs and GDPR export. Installation needs Node.js 18+ with npm 9+, or Python 3.12+; macOS, Windows and Linux are supported. Open source under AGPL v3, with a free local core.
Platforms and languages
SuperLocalMemory Availability
Platforms
Languages
Capabilities
SuperLocalMemory Key Features
Unified local-first memory brain
One local-first memory brain shared by Claude Code, Codex, Cursor, ChatGPT and any MCP/CLI agent, instead of each tool keeping its own isolated memory.
Answer check that never bluffs
Recalled results are graded as supported, related or no answer. Laya judges locally on Apple Silicon Macs; Jev checks online only with explicit consent. It refuses to guess.
Image and document memory
Screenshots, photos and PDFs become searchable locally with Google EmbeddingGemma 2, local OCR, and file-and-page citations on every answer.
Governance and auditability
Roles, scopes, workspaces, save receipts, verifiable full deletion, hash-chained audit logs, GDPR export and an EU AI Act posture report.
Time-aware multi-agent collaboration
Timestamped memories with undoable edits and 'time travel' queries; SLM-Mesh enables multi-machine messaging, locks and shared state for teams.
Best for
Who uses SuperLocalMemory?
Developers running multiple AI agents
Share one governed memory brain across Claude Code, Cursor, Codex and other agents
Plans and access
SuperLocalMemory Pricing
Free
Common questions
SuperLocalMemory FAQs
What is SuperLocalMemory?
A local-first memory system for AI agents: one shared, governed 'brain' that Claude Code, Cursor, VS Code, Windsurf, Codex, ChatGPT and any MCP/CLI agent can read from and write to. Open source under AGPL v3.
Which tools does it work with?
Claude Code, Cursor, VS Code, Windsurf, Codex and more. Run 'slm connect --list' to see the full list of supported integrations.
Does my data stay private?
In the default local mode everything stays on your machine. The Jev online answer-checking step only runs with your explicit consent and an API key.
Can I use it with the web versions of ChatGPT or Claude?
Yes. Connect your GitHub account to link the web versions of ChatGPT and Claude; your database stays on your own machine and each connection gets a free daily quota.
What do I need to install it?
Node.js 18+ with npm 9+, or Python 3.12+. It runs on macOS, Windows and Linux, including Apple Silicon.
Reviews
See what the community thinks and share your experience.
—
Based on 0 ratings
Rating distribution
Community reviews
No reviews
No written reviews yet.
