voiceio vs Superwhisper
Superwhisper is a commercial dictation app for macOS, Windows, iOS and Android that can run Whisper and other models locally or use cloud models, with AI "modes" that reformat what you say. On 2026-10-09 it published a prerelease Linux build that targets Omarchy on x86-64; its docs still list Linux as coming soon. voiceio is free, open-source dictation built for Linux, with a live in-app preview through IBus and fine-tuning on your own voice. Superwhisper has far more model choice and polish across platforms; voiceio is open source and Linux-first.
Which should you use?
Choose Superwhisper if…
- You use macOS, Windows, iPhone or Android, and want one license across them (a lifetime option exists).
- You want a choice of local models (Whisper, Parakeet, Cohere) and cloud models (Deepgram, ElevenLabs, its own S1).
- You want AI modes that turn speech into emails, messages or notes with custom prompts.
- You need meeting recording, file transcription or 100+ languages.
Choose voiceio if…
- You run a Linux desktop other than Omarchy, or want a stable rather than prerelease Linux build.
- You want open source: code you can read, audit and change.
- You want words to appear in the app as you speak; Superwhisper's realtime preview shows in its own window and pastes the result at the end.
- You want a model that learns your voice through fine-tuning.
Feature comparison
| voiceio 1.5 | Superwhisper | |
|---|---|---|
| Platforms | Linux: Wayland and X11 (GNOME, KDE, Hyprland, sway, i3). Untested best-effort path on Windows/macOS | macOS 14+, Windows 10/11, iOS, Android. Linux: prerelease build for Omarchy (x86-64), 2026-10-09 |
| Price | Free, MIT | Free (local Whisper models); Pro $8.49/month, $84.99/year or $249.99 lifetime (docs plans page, 2026-10-10) |
| Where speech is decoded | On your computer. Optional text-only cloud cleanup, off by default | Local or cloud, chosen per mode |
| Speech models | faster-whisper small (default), medium, large-v3-turbo, distil-large-v3; experimental Parakeet and whisper.cpp server | Local Whisper (whisper.cpp), Parakeet, Cohere Transcribe; cloud S1-Voice, Deepgram Nova, ElevenLabs Scribe; LLMs for formatting |
| Open source | Yes (MIT) | No |
| Live typing into apps | Yes: underlined preview in the focused app through IBus, corrected in place, committed when you stop | Realtime preview (Pro, streaming models only) in the recording window; text pasted at the cursor after you stop |
| Custom vocabulary | Hotwords ranked by use (about 35 per decode) plus find-and-replace corrections | Vocabulary hints and replacements (free and Pro) |
| Learns from you | Yes: optional LoRA fine-tune on your own recordings, on CPU, kept only if it wins on held-out clips | Not stated (manual vocabulary, modes, context awareness) |
| Voice commands | Dictation commands only ("new line", "scratch that", "correct that") | Not stated beyond shortcuts and modes |
Superwhisper facts are from its official site, docs or repository, read on 2026-10-10 (sources below). Prices are as listed on that date and can change. voiceio facts are from its README. Spotted something out of date? Open an issue.
Questions
Does Superwhisper work on Linux?
Partly. On 2026-10-09 Superwhisper published a prerelease Linux build that, per its README, "targets x86-64 systems running Omarchy". Its docs still list Linux as coming soon (checked 2026-10-10). voiceio supports GNOME, KDE, Hyprland, sway and i3 on Wayland and X11.
Is there an open-source Superwhisper alternative?
voiceio is MIT licensed and, like Superwhisper's free tier, runs Whisper locally. Voxtype is another open-source option on Linux.
Do both use Whisper?
Yes. Superwhisper's free local models are Whisper via whisper.cpp. voiceio uses Whisper via faster-whisper by default, with experimental Parakeet and whisper.cpp backends.
How much does Superwhisper cost?
Its docs list a free plan with local Whisper models and Pro at $8.49 per month, $84.99 per year or $249.99 once (checked 2026-10-10). voiceio is free.
Try voiceio
Free and MIT licensed. Install it with pipx on Linux, or try the dictation in your browser first: the demo on the home page runs a small speech model inside the tab.