Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking

2026-09-15 · Google DeepMind

Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking

On September 15, 2026, Google introduced two new live dialogue models: Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking. These models are designed to make voice interactions more natural, fluid, and intelligent, capable of handling complex reasoning, real-time visual context, and background task execution without interrupting the conversation.

Model Overview

  • Gemini 3.8 Live: Built for scale and cost efficiency, combining conversational intelligence with fluid dialogue and visual grounding.
  • Gemini 3.8 Live Extended Thinking: Built for high-complexity tasks, with increased intelligence and multi-step reasoning capabilities.

Key Features

Gemini 3.8 Live

  • Near Real-Time Visual Processing: Processes visual inputs in near real-time, enriching conversations with context for more helpful responses.
  • Multilingual Support: Automatically detects and transitions between 97 supported languages mid-conversation.
  • Background Task Execution: Executes tools and API calls in the background while continuing the conversation, allowing the model to acknowledge requests and keep chatting as tasks finish.
  • Use Cases: Guides employee onboarding in real-time, plays chess in near real-time, and builds complete business plans and custom marketing toolkits on the fly through natural speech.

Gemini 3.8 Live Extended Thinking

  • Simultaneous Reasoning and Speaking: Reasons and speaks simultaneously for tasks requiring deeper reasoning, delivering increased intelligence for complex workflows while maintaining an uninterrupted conversational flow.
  • Natural Verbal Cues: Uses early verbal cues like "Let me check that..." to acknowledge prompts naturally.
  • Live Progress Narration: Walks users through multi-step background tasks as they progress.
  • Use Cases: Transforms raw sketches and near real-time voice feedback into functional React components, and coordinates multi-step bookings and asynchronous function calls without interrupting natural live conversation.

Performance

  • Gemini 3.8 Live Extended Thinking:
  • Captured the #1 overall spot on Artificial Analysis' Speech to Speech Quality Index (82.6).
  • Leads in agentic task completion with 68.6% on τ-Voice and 35.1% on Sierra’s τ-Voice-banking benchmark.
  • Scores 97.7% on Big Bench Audio, demonstrating strong reasoning capabilities while maintaining a highly competitive price point.
  • Gemini 3.8 Live:
  • Secured second place in the Speech Agent Arena, showing high user preference.
  • Remains highly cost-effective, providing developers and enterprises with a capable and efficient model built for scale.
  • On ServiceNow’s EVA-Bench, both models push the Pareto Frontier for complex workflows by successfully balancing accuracy with conversational quality.

Availability

These features are available starting today through:

  • Gemini API
  • Google Workspace: Including Docs Live, Gmail Live, and Keep Live.
  • Gemini App
  • Search Live: Provides step-by-step, real-time troubleshooting help powered by Gemini 3.8 Live.

Developer and Enterprise Ecosystem

By using the Gemini Live API, developer platforms enable developers to build and deploy high-performance voice-driven interfaces with ease. Supported platforms include:

  • Agora
  • Fishjam
  • LangChain
  • LiveKit
  • Pipecat
  • Vercel
  • Vision Agents

These platforms manage complex real-time media streams, allowing developers to focus on building application logic and user experiences.

Source