AIGoogle Launches Gemini 3.8 Live and Extended Thinking for Advanced Real-Time Voice AI
Google released Gemini 3.8 Live, optimized for low-latency conversational AI with visual grounding and multi-language support, and Gemini 3.8 Live Extended Thinking, which adds real-time reasoning for complex multi-step tasks.
TLDR
The releases advance multimodal voice and agent capabilities, directly competing with OpenAI's voice offerings. They enable more natural, interruptible, tool-using conversations with external reasoning, signaling rapid iteration in real-time AI and practical progress toward functional AI agents.
Combined views
—
1 Source, first seen 19h ago
Google Launches Gemini 3.8 Live and Extended Thinking for Advanced Real-Time Voice AI
Google released Gemini 3.8 Live, optimized for low-latency conversational AI with visual grounding and multi-language support, and Gemini 3.8 Live Extended Thinking, which adds real-time reasoning for complex multi-step tasks.
TLDR
The releases advance multimodal voice and agent capabilities, directly competing with OpenAI's voice offerings. They enable more natural, interruptible, tool-using conversations with external reasoning, signaling rapid iteration in real-time AI and practical progress toward functional AI agents.