Languages

SayLess transcribes in whatever language you speak. What differs is where that happens, and only one of those two places has a fixed list.

The short answer

Open Settings → Dictation. Under Languages you speak the app shows exactly what your Mac recognises, counted on your machine. If you type a code it cannot do, it says so.

You do not have to set anything. Leave the field empty and SayLess detects the language each time.

Why the list is your Mac's, not ours

The live preview — the words appearing as you speak — comes from Apple's speech recogniser, built into macOS. Its supported set moves with your macOS version and with which languages you have installed, so the honest list is the one your own Mac reports. Publishing a fixed list here would be a promise that quietly goes stale.

That is why the app reads the list at runtime rather than shipping a copy of it.

The on-device engine is a separate model

On Apple silicon, with Produce final text on this Mac turned on, the live preview still comes from Apple but the final transcript comes from a separate on-device model (NVIDIA Parakeet). That model covers European languages. For the others on the list below, turn it off in Settings → Dictation & formatting so Apple's recognition — or your cloud provider — produces the text. Intel Macs never run it, so nothing changes there.

What a Mac on macOS 26 reports today

For reference, not as a guarantee — yours may differ:

Cloud transcription is separate

If you have configured a cloud model, that route is not bounded by the list above; it supports whatever your provider supports. So a language missing from your Mac's list means the on-device live preview cannot do it — not that SayLess cannot transcribe it.

Meeting notes follow the transcript

Notes, decisions and action items are written in the language the meeting was mostly spoken in. Quoted evidence stays verbatim in the transcript's own words, and names stay as they were said.

Last updated 22 September 2026