FAQ
Common questions
What permissions does Okato need?
Permission to capture audio from the app you choose, and the microphone if you want to transcribe yourself or a call. You grant each one once. The audio is processed on your Mac and never leaves it. Screen Recording is only used if you switch on an optional fallback in Settings, and that is off to begin with.
How is Okato different from Apple's Live Captions?
Live Captions is free, built into macOS, and good at captions in the language being spoken. Okato also translates every line into any of 30 languages, Cantonese included. It can listen to one app or a file, labels the speakers, puts subtitles over whatever you're watching, and exports the transcript or burns the subtitles into a new video. If you only ever need captions in one language, Live Captions will do. If you need translation, that's what Okato is for.
Does Okato need an internet connection?
No. Speech recognition, translation and speaker separation all run on your Mac, so it works with the internet off.
What languages can it transcribe?
30 languages, from the world's most spoken to regional ones other apps skip, including Cantonese, which general-purpose speech recognition struggles with.
How does live translation work?
As audio plays, Okato writes it down a line at a time and translates each line straight away. Live mode is built for speed, so the translation arrives with the line rather than a few seconds later.
What is file mode?
Drop a video or audio file into Okato and it takes its time: it transcribes every word, can label who is speaking, and lines it all up with the picture. You can watch the film full screen with subtitles in your language, export the transcript as Markdown, plain text, PDF or timed subtitles, or burn the subtitles into a new copy of the video.
Does it work with subtitles over other apps?
Yes. The subtitles sit over anything on screen, including full-screen video, and clicks go straight through them. You choose the position, size and opacity. There is also a small notepad panel that keeps the conversation so far.
Can it tell speakers apart?
Yes. Okato labels each line with who said it, and you can rename a speaker once for every line, so a meeting or a podcast reads like a script. In File mode you can also move a line to a different speaker.
Can I edit the transcript?
In File mode, yes: correct any line by hand, move it to a different speaker, or remove it. A word you correct is used for the rest of that file, and Okato offers to remember it for every recording in that language. In Live mode the transcript is written as it happens, and you can rename the speakers.
Do I need an account?
No. There is nothing to sign up for, no profile and no email. Okato is a one-time purchase and runs entirely on your Mac.
What Macs does it support?
Okato runs on Apple Silicon Macs, M1 or later, with 8GB of unified memory or more, on macOS 15 (Sequoia) or later.
Is my transcript saved anywhere?
Only on your Mac, and you can delete sessions whenever you like. Export is there if you want a copy somewhere else.