Real-time voice interpretationPowered by Gemini Live

Ready when you are

Language A

Language B

Direction

Audio source

ModelGemini 3.8 Live · Stable · recommended

General live model: translates after each sentence (about 1 s), one session handles both directions. Best for conversations.

No usable key · add one in Settings

00:00Idle

Audio & playback

Input level: —

Play translated audio

Echo cancellation

Speaker feedback guard

100%

Your live interpretation will appear here.

Finished ·In progress