A sophisticated, local-first markdown note-taking application built with React and Rust, Lokus targets users seeking powerful, integrated knowledge management without plugin dependencies. It offers a privacy-focused, extensible platform with built-in advanced features, aiming for superior performance and a streamlined user experience compared to alternatives.

How It Works

Built on Tauri (Rust backend) and React (frontend), Lokus prioritizes speed and a small footprint. Its core differentiator is the "Quantum Performance Architecture," employing techniques like Quantum Superposition Indexing for O(1) lookups, achieving up to 100x faster search. It integrates a native Model Context Protocol (MCP) server for AI assistants and offers VS Code-level extensibility via a robust plugin API. Data remains local, with optional sync via user-chosen cloud providers.

Quick Start & Requirements

• Installation: Pre-built binaries available for macOS, Windows, and Linux (AppImage).
• Build from Source: Requires Node.js v18+, Rust. Clone repo, run npm install, then npm run tauri dev or npm run tauri build.
• Links: Releases: https://github.com/lokus-ai/lokus/releases/latest, Docs: Official Docs

Highlighted Details

• Quantum Search: Benchmarked at 22ms for 10,000 files (109x faster than Obsidian), with 90% less memory usage.
• Built-in Database Views ("Bases"): Transforms markdown/YAML into inline-editable tables, offering Dataview-like functionality natively.
• AI Integration: Native MCP Server provides 68+ tools for AI assistants, supporting local/remote communication and custom plugins.
• VS Code-Level Extensibility: 8 specialized APIs enable deep customization via TypeScript plugins, with WebAssembly and worker thread support.
• Performance: ~10MB application size and <1 second startup time, significantly outperforming Electron-based apps.

Maintenance & Community

Active development with v1.3 ongoing and v1.4 planned for Q1 2026. Community support via GitHub Discussions and Discord. Open to contributions.

Licensing & Compatibility

MIT License: Permissive for personal and commercial use, modification, and distribution.

Limitations & Caveats

Mobile apps are targeted for v1.4 (Q1 2026). Formula support in Bases and E2EE are planned for v1.5 (Q2 2026). While compatible with Obsidian vaults, Obsidian plugins are not directly supported.