lowyap · free beta · Windows x64

Speak. It types. On your computer.

Press a shortcut, talk, and your words appear in any app. Speech recognition runs on your device. Free and open source (MIT).

Download for Windows View source

Beta. Not code-signed yet, so read the install steps below first.

01 — How it works

Three moves. No lowyap account.

01

Press

Hold your shortcut to talk, or tap to start and stop.

02

Speak

Your audio is processed on your device.

03

Release

A local model writes the text where your cursor is.

02 — What you get

Built for people who type all day.

on device

Local speech models

Parakeet, Whisper and others. Downloaded once, then dictation works offline.

optional

Per-app tone

Casual in chat, formal in email. Off by default; needs a local LLM or your own API key.

MIT

Open source

No lowyap account and no telemetry. Read the code or build it yourself.

03 — Offline vs. AI provider

Know what leaves your computer.

Feature Offline Needs
Dictation (speech to text) Yes A downloaded speech model
Per-app tone and AI clean-up Only with a local LLM server A local server such as Ollama, or your own API key. With an API key, the transcript is sent to the provider you pick.
Update checks No GitHub Releases. Can be turned off in settings.
04 — Install safely

Unsigned beta: check, then run.

  1. Download lowyap_<version>_x64-setup.exe and SHA256SUMS from the latest release.
  2. Check the download in PowerShell. The result must match the line in SHA256SUMS. 
    (Get-FileHash .\lowyap_*_x64-setup.exe -Algorithm SHA256).Hash.ToLower()
  3. Run the installer. Windows SmartScreen may say "Windows protected your PC." That means the publisher is unverified, not that a virus was found. If the checksum matched, click More info → Run anyway.

Smart App Control: if it is on in Windows 11, Windows will block lowyap until our signed build is ready. Watch the repository to be notified.

Only download lowyap from its GitHub repository. macOS builds are not distributed yet.