Corrected: Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking

GeminiNew modelConfirmed

Correction: Updated on 4 Oct: Google's article is dated 15 Sep and says rollout started that day. Earlier copy gave only the 17 Sep update date.

Google's new voice models combine conversation, visual context, and background tool use.

Extended Thinking handles more complex reasoning while speaking.

Who can use it: Both models are offered through the Gemini API and AI Studio. Consumer and enterprise access varies by model, app, and subscription.
When: Rollout began September 15, 2026, the article's date; the article was updated September 17. Some enterprise offerings remain in private preview or are coming soon.

Google official announcement Announced Sep 15, 2026, 5:00 PM UTC

Read the official announcement for access details. Availability can depend on your plan, region or workplace settings.

Shared copies of this card may stay cached after a correction. This page is current.