Tencent Meeting Launches AI Interpretation Feature with 3-Second Latency, Voice Cloning Support

According to Beating, Tencent Meeting formally launched its AI interpretation feature today, compressing latency to within 3 seconds and supporting English-Chinese dual-language translation. The system implements real-time translation mechanism, enabling near-simultaneous interpretation and speech.

The feature includes a voice cloning option that mimics the speaker's vocal characteristics, allowing interpreters to sound like the original speaker and helping attendees distinguish between multiple participants. Users can switch back to system-generated voice and adjust the balance between original and interpreted audio.