Google gives Gemini 3.8 Live background thinking

The new voice models deliver simultaneous speech processing and tool execution, featuring SynthID audio watermarking and multi-language auto-detection. Read more…



Source link