Intel / post

Google DeepMind Launches Gemini 3.8 Live and Extended Thinking Models

Source: x.com
Date: Published Sep 15, 2026

Context

Google DeepMind introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, calling them its best conversational AI. It says the models talk, think and handle tasks in the background without breaking the user's flow.

Artificial Analysis later benchmarked them and reports that Extended Thinking, which supports configurable reasoning effort, ranks first at High effort on its Speech to Speech Index and its Tau Voice implementation. The same thread reports slower time to first audio and a higher hourly cost for that variant than for the standard model, so the top score comes with trade-offs.

Terms in this piece · Glossary

benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
tool use — A model's ability to call external functions — run code, search the web, edit files — instead of only generating text.

Read the source x.com