VocalCode vs FluidVoice
Short version first: if you are Mac-only and dictate mostly English, FluidVoice is free, open-source and good — try it with your own microphone and vocabulary. This page exists for the cases where that sentence is not the end of the story. This page is written by VocalCode; factual claims defer to FluidVoice's own repository.
What each one actually is
FluidVoice is on-device macOS dictation, open source under GPLv3, installable with
brew install --cask fluidvoice. Transcription
runs locally, text inserts through the accessibility APIs, and its optional "Fluid Intelligence"
enhancement layer — a privately-maintained local AI runtime — stays on the machine too. When checked on
15 August 2026 it was macOS-only, with Windows and iOS listed as waitlists.
VocalCode is the same local-first philosophy available as a free, open-source app: Windows 10/11 and Apple-silicon macOS today, a dedicated bilingual Chinese–English Paraformer model for Mandarin, Parakeet TDT for European languages, SenseVoice for Korean and Japanese, and the agent-prompting details — acronym collapsing, a user dictionary, a send binding — as the core of the product.
Side by side
| FluidVoice | VocalCode | |
|---|---|---|
| Price | Free | Free and open source |
| Open source | GPLv3 (enhancement runtime private) | AGPL-3.0-only desktop source |
| Where speech is recognised | On your Mac | On your machine |
| Windows | Waitlist (when checked) | Shipping today |
| Mandarin + English code-switching | General model support | Dedicated bilingual Paraformer route |
| Support & accountability | Community | Public issue tracker and support email |
Pick FluidVoice if…
…you are Mac-only, dictate mostly English, and want free and open source. That combination is genuinely excellent, and switching tools may add little to your workflow.
Pick VocalCode if…
- You also work on Windows — both platforms have free installers
- You dictate Mandarin mixed with English identifiers — the bilingual model is the concrete difference
- You want push-to-talk input and local meeting notes in the same app
The honest summary
Both apps are free and open source, with local recognition. Compare platform support, input workflow and model choices using FluidVoice’s source and VocalCode’s source. Also see vs Handy and the Wispr Flow alternative guide.