2026-09-15 · Fuente:Unite AI

Google ships Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, described as its most advanced live dialogue models yet, across Gemini API, AI Studio, Enterprise, Search Live, and Workspace.
Google introduced two new voice models on September 15, 2026: Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking. The pair rolled out simultaneously across a notably wide set of surfaces, spanning the Gemini API, Google AI Studio, Gemini Enterprise, Search Live, Gemini Live, and Google Workspace. That placement puts the same live dialogue capability in front of developers building on the API, enterprises using Google's productivity stack, and users interacting with Search and Gemini Live.
The announcement came from the Gemini Audio Team, presented through a post by Tom Ouyang, a principal engineer, and Malini Jaganathan, a member of technical staff. Google positioned the two releases as its most advanced live dialogue models yet, built for natural interaction. The naming itself — a standard live model paired with an Extended Thinking variant — signals a two-tier structure within the company's voice lineup, separating conversational responsiveness from a deeper-reasoning option at the audio layer.
Distribution breadth is the headline detail for enterprise decision-makers. Launching into Gemini Enterprise and Google Workspace at the same time as developer channels like the API and AI Studio indicates Google is treating live voice dialogue as production infrastructure rather than a limited preview. Adding Search Live and Gemini Live extends the same audio stack into mass-market consumer touchpoints, meaning the models will carry real conversational traffic from day one across both enterprise deployments and consumer-facing surfaces.
For investors tracking the voice-AI segment, the multi-surface launch underscores how central audio dialogue has become to platform strategy. Live dialogue models are engineered for continuous spoken exchange, a workload profile distinct from conventional text inference. With the Gemini Audio Team now shipping numbered iterations at the 3.8 mark, the cadence of live-model releases is one to watch as the conversational AI landscape continues to evolve.
Etiquetas: Google · Gemini 3.8 Live · Tom Ouyang · Malini Jaganathan · Voice AI · Gemini Audio Team
萬安算交所 AIXX · AIXX · Exchange verificado · Ir al mercado