AI voice translator for real-time speech
Xuula translates speech into readable text while it is being spoken — the original on one side, the translation on the other. It is a translator for the conversation you are in, not a phrasebook you type into.
8,000 words free every week. No credit card.
What does the Xuula voice translator do?
Live Translation listens to speech and shows the recognised words beside a translation into a language you choose, updating as people talk. It produces text on screen, not a synthetic spoken voice. The same translation can run beside a meeting transcript, where it is saved with the conversation; standalone Live Translation is on screen only.
Two places translation appears
Live Translation is its own mode: pick a target language, press start, and see what was said next to what it means. Inside Meetings, translation runs beside the live transcript and in movable always-on-top caption windows, and is saved with the conversation when you stop.
What “real-time” means here
Text appears progressively while someone is still speaking and settles as recognition firms up, rather than arriving after the sentence ends. It is fast enough to follow a conversation; it is not an instantaneous simultaneous interpreter, and network conditions affect it.
Text, not a synthetic voice
Xuula does not speak the translation aloud and does not clone or replace anyone’s voice. The output is text you read — which is what makes it usable in a meeting where a second audio track would be in the way.
The original is always kept
Translation is shown beside the source text, never instead of it. In a saved meeting, the original transcript is stored along with the translation, so a disputed phrase can be checked against what was actually said rather than against an interpretation of it.
What it is not
Standalone Live Translation displays results and does not write them to a file or a history — the meeting path is the one that saves. Starting it stops dictation, and it will not start while a meeting recording is running. And Voice Typing never translates: dictated text is always written in the language you spoke.
Voice translation questions
It is live and progressive: text appears while a person is still talking and settles as recognition improves. It is not instantaneous, and it depends on your connection — treat it as following a conversation, not as a simultaneous interpreter.
No. The output is text on screen — a source pane and a translation pane, or caption windows over a meeting. There is no synthesized speech and no voice cloning.
Translation that runs beside a meeting is saved with that conversation on your PC. Standalone Live Translation is displayed only: it is not written to a file or added to your history.
No. Voice Typing writes what you said, in the language you said it. Translation is a separate mode with its own screen — you can read and copy the translated text, but dictation will not silently type another language into your document.
The languages the app names for speech — English, Russian, Spanish, French, German, Italian, Portuguese, Ukrainian, Polish, Chinese, Japanese, Korean, Arabic and Hindi — with a target language you pick. Quality varies by pair, so test the pair you need on the free plan.