Gemini News 09/15/2026 AI Rating: High

Google Launches Gemini 3.8 Live and 3.8 Live Extended Thinking for Real-Time Voice

#Gemini#Google#GeminiLive#VoiceAI#RealTime

Google launched two new real-time voice models: Gemini 3.8 Live, built for cost-efficient scale, and Gemini 3.8 Live Extended Thinking, aimed at complex, multi-step reasoning tasks conducted through live dialogue.

Details

  • Gemini 3.8 Live: near real-time visual processing with contextual awareness, support for 97 languages with automatic mid-conversation switching, and background tool and API execution that doesn’t interrupt the dialogue
  • Extended Thinking adds: simultaneous reasoning and speech for complex workflows, deeper analytical capabilities for enterprise tasks, and natural verbal acknowledgments like “Let me check that…” with live progress narration during multi-step operations
  • Benchmarks: Extended Thinking scores 82.6 on Artificial Analysis’ Speech to Speech Quality Index, 68.6% on τ-Voice, and 97.7% on Big Bench Audio reasoning tasks; standard Live ranks second on Speech Agent Arena while leading on cost-effectiveness
  • Availability: both launched September 15, 2026 via the Gemini API and AI Studio for developers, Gemini Enterprise in private preview for organizations, and consumer products including Search Live, Gemini Live, Workspace, Gmail, and Keep

What happened next

Splitting the Live lineup into a cost-optimized tier and a reasoning-focused Extended Thinking tier mirrors the text/flash-vs-reasoning split Google has used elsewhere in its model lineup, applied for the first time to real-time voice — giving developers a choice between cheap, fast voice interaction and slower, deeper reasoning delivered conversationally.

Source https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-gemini-3-8-live-extended-thinking/

Related Articles

More Gemini articles →
Gemini News 10/01/2026 Read AI Rating: Medium

Google Launches Guided Vision in Gemini Live, a Real-Time Accessibility Feature for Blind and Low-Vision Users

Google introduced Guided Vision in Gemini Live, a voice-forward accessibility feature that analyzes a live camera feed and gives real-time verbal guidance, built with Aira and over 1,000 blind and low-vision testers.
Gemini News 09/24/2026 Read AI Rating: Medium

Google Adds Live Avatar to Gemini 3.8 Live, an Animated Video Layer for Real-Time Conversation

Google launched Live Avatar for Gemini 3.8 Live, pairing near real-time video generation with speech for lip-synced, expressive animated conversations across 97 languages, currently exclusive to Gemini Enterprise with SynthID watermarking on all output.
Gemini News 09/23/2026 Read AI Rating: High

Google Launches Gemini 3.8 Flash and Flash-Lite TTS With 2,000+ Voices

Google introduced Gemini 3.8 Flash TTS and Flash-Lite TTS, its most expressive audio generation models yet, supporting 100+ languages, voice creation from text prompts, cloning from 30-second samples, and long-form multi-speaker generation, topping Hume AI's voice quality benchmarks.
Gemini News 08/26/2026 Read AI Rating: Medium

Gemini Live Adds Voice-Driven Multi-Step Task Automation

Google added new productivity features to Gemini Live — Spark for multi-step tasks across Docs, Sheets, and Drive, a spoken Daily Brief combining Gmail and Calendar, hands-free inbox sorting, and Personal Intelligence — all controlled by voice.