Simultaneous translation for macOS and Windows
Callisto listens to your computer audio and translates while people speak: voice in your headphones, captions on screen. In meetings, videos and live classes, with no bot joining the call and no switching apps.
01 How it works
Callisto sits between your computer sound and your ears. Nothing joins the meeting, and no one else has to install anything.
System audio (what you are hearing) or your microphone, captured right on your computer.
Recognizes speech and translates it as a stream, phrase by phrase, into one of 9 languages.
Translated voice in your headphones and captions on screen. Captions arrive first, the voice confirms.
02 Modes
Translates what you hear and dubs it into your headphones. Unlike the built-in translation in Meet or Teams, it works in any app, including meetings where you are just a guest.
You hear the meeting in your language and speak in your language. The virtual microphone delivers your voice, already translated, to everyone else.
03 Spec sheet
04 Pricing
Human interpreting starts at about R$ 5 per minute. Callisto starts at R$ 0.40.
Purchases happen inside the app, by Pix or card. Credits never expire.
05 Download
The first 15 minutes are on us. After that, buy minutes in the app whenever you need them.
14.2 or later · Apple Silicon and Intel
Open the file and drag Callisto into Applications. The first time, macOS asks for confirmation: in System Settings → Privacy & Security, click Open Anyway.
10 and 11 · 64-bit
If Windows warns about an unrecognized app, click More info and then Run anyway.
06 FAQ
Still have a question? Write to oi@usecallisto.app.
Captions appear first, and the translated voice arrives on average about 3 seconds after the speaker. It varies with the sentence: the model sometimes waits for the end of an idea before translating. That is the same range simultaneous interpreters work in (2 to 4 seconds).
No. Audio is processed as a stream and discarded. We keep no recordings; transcripts only exist if you turn on history. Details in the privacy policy.
In Brazil, a participant may process their own conversation audio. Still, telling people is good practice, and in other countries it may be required. You are responsible for how you use it, as with any recorder or note-taking assistant.
In Listen mode the dub plays alongside the original audio. With speakers, the dub would be captured again and turn into an echo. With headphones, that does not happen.
Through the virtual microphone: the app installs it in one click on macOS, and you pick it as the microphone in Zoom, Meet or Teams. Your speech comes out translated, in a synthesized voice, in the meeting language.
No. A minute you buy is yours until you use it.
Yes. This first version is not yet signed through Apple and Microsoft, so the system asks you to confirm. On macOS, open System Settings → Privacy & Security and click Open Anyway. On Windows, click More info and then Run anyway.