Gemini 3.8 Live is a Large Language Models (LLMs) tool. Voice AI models for natural conversations with background reasoning and task execution. Key features include Native Speech-to-Speech Processing, Background Task Execution, and Extended Thinking Mode. Best for software developers and engineers, customer service representatives and content creators.
About Gemini 3.8 Live
Key Features
<strong>Native Speech-to-Speech Processing.</strong> Gemini 3.8 Live processes audio directly without needing separate transcription or text-to-speech steps, creating more natural voice interactions.
<strong>Background Task Execution.</strong> The model runs tool calls and API requests in the background while continuing the conversation, so users don't experience awkward pauses or interruptions.
<strong>Extended Thinking Mode.</strong> Gemini 3.8 Live Extended Thinking adds multi-step reasoning for complex tasks like debugging or programming tutoring, ranking first on speech quality benchmarks with an 82.6 score.
<strong>97-Language Support.</strong> Both models automatically detect and switch between 97 languages mid-conversation, making them practical for global customer service and multilingual workflows.
<strong>Real-Time Visual Grounding.</strong> The models can process visual input in near real time during voice conversations, allowing users to discuss images or sketches while speaking.
<strong>Cost-Effective Pricing.</strong> Gemini 3.8 Live is priced at $0.005 per minute for audio input and $0.018 per minute for audio output, making it accessible for high-volume voice applications.
Frequently Asked Questions
Gemini 3.8 Live is a real-time voice AI model from Google DeepMind that enables natural conversations with AI. It processes speech directly and can handle tasks in the background while maintaining dialogue. There are two versions: the standard model for cost-effective, high-volume use, and Extended Thinking for complex reasoning tasks.
Gemini 3.8 Live costs $0.005 per minute for audio input and $0.018 per minute for audio output through the Live API. This pricing applies to both the standard and Extended Thinking versions, making it competitive for production voice applications.
Gemini 3.8 Live Extended Thinking is built for high-complexity tasks that need multi-step reasoning. It ranks first on Artificial Analysis' Speech to Speech Quality Index with a score of 82.6 and provides spoken status updates during longer tasks. The standard version focuses on cost efficiency and scale for simpler voice interactions.
You can access Gemini 3.8 Live through the Gemini API, Google AI Studio, and Gemini Enterprise. It's also available in Google Workspace apps like Docs, Gmail, and Keep for Pro and Ultra subscribers, as well as in the Gemini Live app and Search Live.
