Lexena

AI-powered voice dictation, notes, and personal vocabulary — for your Windows desktop.

Lexena 3.0 is in public Beta. The core experience is stable; we are polishing things and shipping new capabilities frequently.

Highlights

• 🎤 Push-to-talk dictation with configurable global hotkeys, instant insertion into any active window.
• 📝 Built-in notes — rich text editor with folders, backlinks, code blocks, tables, and task lists.
• 📊 Personal dashboard — streaks, time saved, top words, activity heatmap.
• 👤 Multi-profile — fully isolated workspaces per profile.
• ☁️ Optional account & sync — preferences, dictionary, and snippets follow you across devices.
• 🔒 Offline or cloud transcription — your choice.
• 📚 Personal dictionary & snippets — improve accuracy and expand text on the fly.
• 🪟 Floating mini-window, system-tray integration, and auto-paste.
• 🔄 Signed auto-updates and a 🌍 EN / FR interface.

Get Lexena

Head to the Releases page and grab the installer that fits you best — NSIS (recommended), MSI, or portable. Launch it, and you are ready to dictate.

Sign in to enable sync across your devices, or stay fully local — your choice. Lexena currently runs on Windows only.

A quick tour

Notes — a full editor next to your transcriptions

History — every dictation, searchable, exportable

What you can do

Dictate anywhere. Bind a hotkey, talk into your microphone, and Lexena pastes the transcription straight into the active window — chat, email, code editor, browser.

Capture and organise. Keep notes alongside your transcriptions in the built-in editor: folders, backlinks, tables, code blocks, task lists.

Make it yours. Tune your personal dictionary for tricky words, define snippets that expand into longer text, customise your hotkeys, theme, and run multiple profiles for different contexts.

Beta status

Lexena 3.0 is a public Beta — usable every day, but moving fast. New features land regularly and a few rough edges are still being smoothed out.

Found a bug or want to suggest something? Open an issue — feedback shapes the roadmap.

Build from source

For contributors and curious developers.

Requirements

• Node.js (LTS) and pnpm
• Rust toolchain via rustup
• Visual Studio Build Tools with the Desktop development with C++ workload
• LLVM and CMake in your PATH — required by whisper-rs to build the native Whisper backend (see the Build Requirements notes in CLAUDE.md)

Run and build

pnpm install
pnpm tauri dev      # dev (frontend + Rust backend)
pnpm tauri build    # signed production build

If the build fails on cmake / MAX_PATH errors, disable the Vulkan GPU backend and fall back to CPU:

pnpm tauri build -- --no-default-features

Release process

Releases are orchestrated by a single script:

.\scripts\make-release.ps1 -Version 3.x.x          # stable
.\scripts\make-release.ps1 -Version 3.x.x-beta.N -Beta   # beta

Update signing keys live in GitHub Secrets — see docs/UPDATER_SETUP.md for the full setup.

Tech stack

Layer	Technology
Frontend	React 19, TypeScript, Tailwind v4, Vite, Tiptap
Backend	Rust, Tauri v2, cpal, enigo, whisper-rs
Sync	Supabase (Auth + Postgres + Edge Functions)
Distribution	GitHub Releases — NSIS / MSI / portable, signed updates

Links

• 🌐 Website: lexena.app (coming soon)
• 📦 Releases
• 🐛 Issues

License

Released under the MIT License.