Changelog

What's new

Every release, straight from GitHub Releases.

1. v1.0.0

   LatestOctober 22, 2026

   Abe 1.0.0

   Abe 1.0.0

   Free, open-source voice-to-text for your desktop. Hold a key, talk, and your words appear wherever you’re typing.

   This is the first open-source release of Abe. It replaces the old hosted service: there are no accounts, subscriptions or servers.

   Highlights

   • Two transcription engines
     • Local (offline): Whisper runs on your computer. Choose the tiny, base, small or large-v3-turbo model. It’s free, works offline, and your audio never leaves your machine.
     • Your own API key: Groq (recommended: fast and cheap), OpenAI, or any OpenAI-compatible endpoint. Audio goes straight from your computer to your provider, and keys are stored in your OS keychain.
   • Voice Transform: select text, hold Ctrl+Alt+T, and say “fix grammar”, “make it formal” or “translate to Spanish”.
   • Dictate anywhere: push-to-talk or toggle with Ctrl+Alt+Space. Paste your last transcription again with Ctrl+Alt+V.
   • Word dictionary, local history, usage stats with estimated provider cost, audio ducking, and typewriter output mode.
   • Private by design: no telemetry, no analytics, no Abe servers.

   Download

   Platform	File
   Windows 10/11 (x64)	Abe-windows-x64-setup.exe (signed)

   macOS and Linux builds are coming soon. You can also build from source.

   Getting started

   1. Run the installer and open Abe.
   2. Choose Local (offline) or Use my own API key. New to API keys? Follow the Groq guide.
   3. Hold Ctrl+Alt+Space and talk.

   Docs: https://getabe.app/docs · Support Abe: https://buy.stripe.com/dRm28q7VlgEa8KZ5Rq3Je00

   Licensed AGPL-3.0-or-later.