Alphabet Inc. Class C (GOOGL.US) releases Gemini 3.8 Live dual models: voice AI enters the era of "reasoning while conversing," with benchmark scores surpassing OpenAI and SpaceXAI.

Alphabet Inc. Class C (GOOGL.US) released two new Gemini models on TuesdayGemini 3.8 Live and Gemini 3.8 Live Extended Thinkingfocused on improving voice agent performance. Alphabet Inc. Class C engineer Tom Ouyang said in a blog post: "For developers and enterprises, these models provide the foundational building blocks for building reliable, production-ready voice agents. They also make conversations with Gemini more fluid and collaborative for users through the Gemini app, Google Workspace, and Searchhelping users complete complex tasks using voice alone." It is understood that Gemini 3.8 Live Extended Thinking scored 82.6% on the Artificial Intelligence Speech to Speech Index, narrowly surpassing OpenAI's GPT-Live-1 Astra (81.5%) and SpaceXAI's (SPCX) Grok Voice Think Fast 2.0 (81.3%). Ouyang said: "For tasks requiring deep reasoning, 3.8 Live Extended Thinking can reason while speaking. It brings a higher level of intelligence to complex workflows while maintaining conversational coherence and fluidityresponding naturally to users through early voice cues such as 'Let me look that up...' and guiding users through the progress of multi-step background tasks with real-time progress narration." In addition, Alphabet Inc. Class C has partnered with Salesforce, Inc. (CRM.US), Genspark, and Lumeris to explore application scenarios for the new models. Bob Van Osten, Vice President of Agentforce products at Salesforce, Inc., said: "By combining real-time multimodal capabilities with agentic AI, we are exploring new ways to create richer, more intuitive customer experiences, from first contact to problem resolution." According to the introduction, Gemini 3.8 Live is available to developers starting today through the Gemini API and Google AI Studio, with a private preview in Gemini Enterprise, and is launching for all users in Search Live. Gemini 3.8 Live Extended Thinking is available to developers starting today through the Gemini API and Google AI Studio, with a private preview in the enterprise version Gemini Enterprise, and is launching for all users in Gemini Live and Google AI Pro.