Gemini 3.8 Live Extended Thinking powers Gemini Live, Gmail, & Keep

Google has unveiled Gemini 3.8 Live and its extended version, Gemini 3.8 Live Extended Thinking, touting them as the company's most sophisticated live dialogue models to date. Building upon the success of the March debut of 3.1 Flash Live, the latest models strive to elevate user interaction with artificial intelligence to new heights of intuitiveness and intelligence.

As suggested by their names, Gemini 3.8 Live Extended Thinking provides heightened intelligence and the ability to tackle multi-step reasoning. This makes it particularly well-suited for intricate tasks, such as the recent introduction of Gmail Live (a conversational search feature), Docs Live (for drafting and editing), and Keep Live (for note creation). The model is also set to roll out alongside Gemini Live.

Gemini 3.8 Live Extended Thinking "reasons and speaks simultaneously," offering increased intelligence for complex workflows while preserving a seamless conversational flow. It uses early verbal cues like "Let me check that..." to naturally acknowledge prompts, and live progress narration to guide users through multi-step background tasks as they advance.

In contrast, the Gemini 3.8 Live base model emphasizes "conversational intelligence with fluid dialogue and visual grounding." This model can process visual inputs in near real-time, with tools and API calls executing in the background while the conversation continues. Users can anticipate the model to "acknowledge requests and keep chatting while tasks finish in the background."

Furthermore, it can seamlessly detect and switch between 97 supported languages mid-conversation. Gemini 3.8 Live is designed for scalability and cost efficiency, boasting a commendable "second place in the Speech Agent Arena." It will drive AI Mode's Search Live experience. For more updates, tune into 9to5Google on YouTube.