The Local Dictation Setup Guide — Free, Open-Source, 100% Offline Dictation for Mac, Windows & Linux

The companion to the full video. Stop paying $15/month to talk to your computer: this is the complete setup manual for free, open-source, fully offline dictation on every OS — with every star count verified against the GitHub API on July 23, 2026. INSIDE: (1) THE DECISION CARD — nine tools matched to nine needs on one page: Handy (27,238 stars, MIT, Win/Mac/Linux, zero cloud) for 'it just works', OpenWhispr (MIT, cross-platform) for AI cleanup + meeting notes with speaker labels, FluidVoice for maximum Mac speed, VoiceInk ($25 prebuilt or build free — yes, it IS open source), Hex for zero settings, Voxtype (Wayland-native, 7 engines, every GPU vendor) and Speech Note for Linux, Vibe for transcribing files, Whispering for the browser. (2) INSTALL PAGES PER OS — the exact brew / winget / flatpak / AUR commands, plus the permissions and hotkey setup. (3) THE MODEL PICKER — Whisper vs NVIDIA Parakeet by machine, and why Parakeet (~3,380x real-time, tops the English leaderboard, runs on a plain CPU, CC-BY licensed) is the reason all these apps suddenly feel instant. (4) THE WATCH LIST — the popular 'free' apps that are NOT open source (Spokenly says so on its own site, MacWhisper, Wispr Flow, superwhisper) plus the honest catches on the open picks, including FluidVoice's private Fluid-1 formatting model. (5) DO THIS TONIGHT — the 5-step, 10-minute path from install command to dictating into your AI chat box, where speaking a 200-word prompt takes 1 minute instead of 4. Independent field manual from Hyperautomation Labs — not affiliated with any of the projects.

By submitting, you'll be subscribed to the Hyperautomation AI Report newsletter — one practical AI email per week. Unsubscribe anytime.

Free. No spam. Unsubscribe anytime.