voiceio vs Aqua Voice
Aqua Voice is a polished commercial dictation app for Mac, Windows, iPhone and Android. It runs its own speech model, Avalon, in the cloud, adds language-model formatting, and can read on-screen context to get names and code right. voiceio is free, open-source dictation for Linux that decodes speech on your own computer and fine-tunes itself on your voice. Aqua does not run on Linux and needs a connection for dictation; voiceio runs only on Linux, and does not try to match Aqua's AI editing features.
Which should you use?
Choose Aqua Voice if…
- You use macOS, Windows, iPhone or Android, and want one subscription across them.
- You want AI formatting and editing: custom instructions, Edit Mode (rewrite selected text by voice), Deep Context from what is on screen.
- You want a managed product with team billing, SSO, SOC 2 Type II and ISO 27001.
- You dictate in many languages (Aqua lists 49).
Choose voiceio if…
- You are on Linux. Aqua's own comparison page says it does not run there.
- You want audio to stay on your machine and dictation to work offline.
- You want software you can read and change (MIT), with no account and no word limits.
- You want the model itself to learn your names from your own voice, not only a word list.
Feature comparison
| voiceio 1.5 | Aqua Voice | |
|---|---|---|
| Platforms | Linux: Wayland and X11 (GNOME, KDE, Hyprland, sway, i3). Untested best-effort path on Windows/macOS | Mac, Windows 10/11, iOS 17+, Android 10+. Not Linux (a CLI for settings and file transcription installs on Linux) |
| Price | Free, MIT | Free (1,000 words); Pro $8/mo yearly or $10 monthly; Max $24/mo yearly or $30 monthly; Business from $19/seat/mo (pricing page, 2026-10-10) |
| Where speech is decoded | On your computer. Optional text-only cloud cleanup, off by default | Cloud; internet needed for cloud dictation |
| Speech models | faster-whisper small (default), medium, large-v3-turbo, distil-large-v3; experimental Parakeet and whisper.cpp server | Own model, Avalon, plus language-model post-processing; Realtime mode uses a third-party provider |
| 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 | Default inserts after you stop. Realtime Mode (Max plan) shows words live in Aqua's toolbar; text lands in the app when you finish |
| Custom vocabulary | Hotwords ranked by use (about 35 per decode) plus find-and-replace corrections | Dictionary, replacements, custom instructions, synced across devices |
| 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 dictionary; optional screen context) |
| Voice commands | Dictation commands only ("new line", "scratch that", "correct that") | Edit Mode, "Send it" (Max), Computer Control beta on macOS |
Aqua Voice 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 Aqua Voice work on Linux?
No. Aqua's own comparison page says it runs on Mac, Windows, iPhone and Android (checked 2026-10-10). A separate Aqua CLI installs on Linux, but it manages settings and transcribes files rather than providing live dictation.
Is there an offline Aqua Voice alternative?
voiceio runs its speech model on your computer and works offline after a one-time model download. On Linux, Voxtype and nerd-dictation are also local.
Is Aqua Voice more accurate than Whisper?
Aqua publishes benchmarks for its Avalon model, including Open ASR Leaderboard results. voiceio's approach is different: fine-tune a local Whisper model on your own recordings and measure it on your held-out clips with voiceio learn eval.
How much does Aqua Voice cost?
As listed on aquavoice.com/pricing on 2026-10-10: a free tier with 1,000 words, Pro at $8/month billed yearly ($10 monthly), Max at $24/month billed yearly ($30 monthly), and Business from $19 per seat per month billed yearly. 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.