Google introduces Gemini 3.8 Live and Live Extended Thinking
One user reports testing Gemini 3.8 Live on three books with a phone camera and voice, asking for explanations and a bedtime-story version of a page—without PDFs or uploads.
TLDR
Google announced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, calling them its most advanced audio models yet. The company says they are built for natural, production-ready voice applications, including collaboration and complex tasks handled by voice.
A user testing Gemini 3.8 Live said they pointed a phone camera at three books: asking for a Sanskrit verse from the Bhagavad Gita to be explained for a Gen Z developer in their 20s, a quick explanation of an Atomic Habits diagram, and a warm bedtime-story version of a Chicken Soup for the Soul page. They reported using no PDFs or uploads and said the model understood the context better than they expected.
Combined views
6.3K
3 Sources, first seen ago