Tencent Meeting just rolled out an AI simultaneous interpretation feature that can not only translate in real time with under three seconds of delay but also mimic the speaker's voice. Yes, your AI translator might soon sound just like you — at least in another language.
The update supports Chinese-English translation for now. Instead of waiting for a full sentence like a traditional relay, Tencent Meeting's AI processes speech as it happens, keeping translation nearly in sync with the speaker. In multi-person meetings, enabling voice mimicry means each translated voice retains the speaker's unique characteristics, making it easier to tell who's saying what — no more robotic monotones. Users can switch back to system voices at any time and adjust the volume balance between original audio and interpretation.
This new interpretation service integrates tightly with Tencent Meeting’s existing real-time transcription and in-meeting captions. Together, they create a full AI-powered workflow: bilingual subtitles, text records, and interpreted audio. This closed loop captures and passes along cross-language speech instantly, generating high-quality context that improves AI meeting notes, action item extraction, and post-meeting Q&A accuracy.