| I wanted Whisper-level dictation on my Intel 2020 iMac without: * paying a monthly fee for the privilege of using my own API key, * hauling an 800 MB local model, or * copy-paste gymnastics from a floating window. So I built VTS (Voice Typing Studio). What it is: * 5 MB notarised menu-bar app. Press a toggle hot-key → talk → text lands wherever the cursor is. * Bring-your-own OpenAI / Groq key. Nothing touches my servers. * Universal binary – runs on Intel and Apple silicon. * MIT-licensed; core stays free/open. If future features justify it, a pro tier may appear. Repo, DMG, and 8-second demo video:
https://github.com/j05u3/VTS Why it’s different Previous HN tools solve pieces of the problem, but not all at once: * superwhisper – fully offline, but big model and Apple-silicon only. * MacWhisper – accurate, yet still a separate window/copy-paste flow. VTS aims for “menu-bar switch and forget.” Would love feedback on onboarding, features or anything that breaks. |