AnalysisAI ModelsSeptember 25, 2026

Google DeepMind and Sierra AI discuss native audio models for real-time voice

Read original source →youtube.com

Valeria Wu (Google DeepMind) and Soham Ray (Sierra AI) cover conversational latency, multilingual code-switching, and agentic voice tasks in a Google for Developers session. They also address how new benchmarks can better evaluate real-time voice experiences.

People · Valeria Wu, Soham Ray

2 sources

More stories today

Open the live feed