A native macOS development environment built for developers who want their tools to get out of the way. Multi-language, wearables-aware, AI-native — and crafted to feel like an extension of how you think.
No bloat, no telemetry, no popups asking you to upgrade. Just a fast editor that respects your attention and gets sharper the more you use it.
Built with a Rust engine and WebView host (`tao` + `wry`). Zero React/Vite in the shipped path. Sub-second cold starts with JSON IPC bridging the native OS to the vanilla TypeScript shell.
A lightning-fast ⌘P palette for rapid navigation. A full-page, searchable settings view configured entirely via JSON.
A real AI Agent panel with streaming transcripts and MCP-backed flows. Deep integration with local LLMs (vLLM, Ollama, LM Studio) or cloud endpoints via ⌘L.
LSP diagnostics mapped natively to the UI. Live problem counts in the status bar and inline error overlays.
Vim mode, Outline panel, and an intuitive breadcrumb navigation. Configurable to your bone with sane defaults.
No telemetry. Code never leaves your machine unless you explicitly invoke cloud AI. Sync config securely via iCloud.
Tupã itself is a showcase of what Lx8 Labs ships: native, low-latency, no-compromise.
Tupã is in active Phase 1.2 development. Scroll right to see the feature pipeline.
Buffer, undo tree, multi-cursor, Tree-sitter, generic LSP host. Goto-definition, diagnostics.
Global Command Palette (⌘P), Settings full page, and synchronized Outline Panel for large files.
Real-time sending and streaming transcript with agent endpoints (vLLM, Ollama, LM Studio, MLX).
Advanced workspace navigation, deepened diagnostic integrations, and enhanced multi-window workflows.
Proprietary high-performance text rendering pipeline for an ultra-low latency typing experience.
Apple Watch app for build status and breakpoint nudges. Vision Pro multi-window editor surface.
Join the waitlist for the public beta. We'll email when invites go out — no spam, no marketing drips.