AiThaoWhat it does for you
- Each person speaks and hears their own language during the call.
- Where a language has no spoken translation, the other person reads subtitles instead.
- Until a translation arrives, you hear the other person’s own voice rather than silence.
How it works
When both people agree, a translation worker joins the call as a silent participant. It takes each person’s audio, cuts it into short chunks and streams them to a cloud speech model, which recognises the speech, translates it and speaks the result. The worker plays the translated speech into the call for the other person, or sends subtitles for languages without a voice.
Translation workers run as separate processes from the rest of the server, so many translated calls at once do not slow each other or ordinary chats down.
What affects the result
- A steady internet connection on both sides.
- One person speaking at a time, in short sentences.
- A quiet place: background noise makes recognition worse.
Under the hood
WebRTC · streaming speech recognition · machine translation · speech synthesis
Questions
Is my call recorded to be translated?
No. Audio is streamed to the translation service while you speak, and AiThao does not record the call.
Why is there a short delay?
Speech has to be recognised, translated and spoken again. The delay depends on the network, the languages and how long each sentence is.