SpeedOfTalk vs FluidVoice
FluidVoice is a free, open-source dictation app for the Mac, and a popular one. One of the speech models it offers is the same one SpeedOfTalk uses. Here's how the two differ, so you can pick the one that fits.
| SpeedOfTalk | FluidVoice | |
|---|---|---|
| Price | $9.95 once | Free, no paid tiers source |
| Source code | Closed | Open source (GPLv3). Its optional local cleanup model is kept private. source |
| Where speech is processed | On your Mac | On your Mac source |
| Cleanup | On your Mac. Smart cleanup needs Apple Intelligence turned on. | Its own local model (about 3.5 GB extra), or a cloud provider you choose, such as OpenAI or Groq source |
| Usage analytics | None | Anonymous, on by default; can be turned off, after which a weekly activity signal is still sent source |
| Speech models | One, chosen and tested for you | Several to choose from source |
| Languages | 25 European languages | Up to 99, depending on the model source |
| Macs | macOS 26 Tahoe or later, Apple silicon | macOS 15 or later; Intel Macs too, with some models source |
| Other platforms | None | Windows and iPhone in beta source |
| Help | From us, through the contact form, with a 14-day refund | Community: GitHub and Discord source |
FluidVoice details from fluidvoice.com and its GitHub page (version 1.6.11). Checked October 2026. Free apps change fast; check their page for the current state.
Choose SpeedOfTalk if…
- You want one setup that's been chosen and tested for you, rather than picking models and providers.
- You want no analytics at all, and nothing that can send your words anywhere.
- You'd like someone to answer when something breaks, and your money back if it doesn't suit you.
- You'd rather not download a 3.5 GB cleanup model.
Choose FluidVoice if…
- You don't want to pay. It's free, and good.
- You want to read or change the source code.
- Your Mac has an Intel chip, or is still on macOS 15 Sequoia.
- You dictate in a language outside Europe.
- You also want it on Windows, or want to pick your own models and AI providers.