[{"data":1,"prerenderedAt":30},["ShallowReactive",2],{"nr-en-google-gemini-3-8-live-sprach-ki":3},{"slug":4,"title":5,"dek":6,"date":7,"time":8,"publishedAt":9,"updated":10,"updatedAt":10,"dateFmt":11,"updatedFmt":10,"kind":12,"tier":13,"author":14,"authorName":15,"topics":16,"tracker":22,"trackerLabel":23,"headlineStat":24,"image":25,"ogImage":26,"imageAlt":5,"csv":10,"minutes":27,"words":28,"html":29},"google-gemini-3-8-live-sprach-ki","Google Launches Gemini 3.8 Live – Voice AI with Real-Time Reasoning","Google has released two new audio models that equip voice agents with advanced reasoning capabilities. Gemini 3.8 Live Extended Thinking tops benchmarks and positions itself as a competitor to OpenAI's GPT-Live-1.","2026-09-16","08:14","2026-09-16T08:14:00+02:00","","September 16, 2026","news","standard","ideal-syka","Ideal Syka",[17,18,19,20,21],"Voice AI","Gemini","Google","Extended Thinking","Voice Agents","\u002Fstand-der-ki","AI Progress","Gemini 3.8 Live Extended Thinking leads Speech-to-Speech Quality Index with 82.6 points","\u002Fnewsroom\u002Fimg\u002Fgoogle-gemini-3-8-live-sprach-ki.webp","\u002Fog-nr\u002Fgoogle-gemini-3-8-live-sprach-ki.en.png",2,425,"\u003Cp>Google unveiled \u003Cstrong>Gemini 3.8 Live\u003C\u002Fstrong> and \u003Cstrong>Gemini 3.8 Live Extended Thinking\u003C\u002Fstrong> on September 15, 2026 – two new voice AI models designed to enable more natural and intelligent speech interactions. The models can handle complex tasks in the background without interrupting conversations, setting a new standard for voice-powered AI agents.\u003C\u002Fp>\n\u003Ch2>The essentials\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Cstrong>Gemini 3.8 Live Extended Thinking\u003C\u002Fstrong> leads Artificial Analysis' Speech-to-Speech Quality Index with \u003Cstrong>82.6 points\u003C\u002Fstrong>\u003C\u002Fli>\n\u003Cli>Model achieves \u003Cstrong>68.6% on τ-Voice\u003C\u002Fstrong> and \u003Cstrong>35.1% on Sierra's τ-Voice-banking benchmark\u003C\u002Fstrong> for agentic task completion\u003C\u002Fli>\n\u003Cli>Available immediately via \u003Cstrong>Gemini API, Google Workspace, and the Gemini app\u003C\u002Fstrong>\u003C\u002Fli>\n\u003Cli>Led by \u003Cstrong>Tom Ouyang\u003C\u002Fstrong> (Principal Engineer) and \u003Cstrong>Malini Jaganathan\u003C\u002Fstrong> (Technical Staff) of the Gemini Audio Team\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2>Two models for different use cases\u003C\u002Fh2>\n\u003Cp>Google differentiates between two variants: \u003Cstrong>Gemini 3.8 Live\u003C\u002Fstrong> is designed for scalability and cost efficiency, combining conversational intelligence with fluid dialogue and visual grounding. \u003Cstrong>Gemini 3.8 Live Extended Thinking\u003C\u002Fstrong> targets complex tasks with increased intelligence and multi-step reasoning.\u003C\u002Fp>\n\u003Cp>Both models can \u003Cstrong>handle interruptions, switch between languages, and explain their reasoning process\u003C\u002Fstrong> while working. This approach aims to make AI interaction more intuitive and less frustrating for users.\u003C\u002Fp>\n\u003Ch2>Benchmark leadership and performance\u003C\u002Fh2>\n\u003Cp>Gemini 3.8 Live Extended Thinking delivers strong benchmark results: the model scores \u003Cstrong>97.7% on Big Bench Audio\u003C\u002Fstrong> and leads multiple industry benchmarks. These numbers suggest Google is currently at the forefront of voice AI quality – particularly significant when compared to OpenAI's \u003Cstrong>GPT-Live-1\u003C\u002Fstrong>, which is positioned as a competing product.\u003C\u002Fp>\n\u003Cdiv class=\"tbl-scroll\">\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Metric\u003C\u002Fth>\n\u003Cth>Score\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>Speech-to-Speech Quality Index\u003C\u002Ftd>\n\u003Ctd>82.6\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>τ-Voice (Agentic Task Completion)\u003C\u002Ftd>\n\u003Ctd>68.6%\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Sierra's τ-Voice-banking\u003C\u002Ftd>\n\u003Ctd>35.1%\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Big Bench Audio\u003C\u002Ftd>\n\u003Ctd>97.7%\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\u003C\u002Fdiv>\n\u003Ch2>Availability and access\u003C\u002Fh2>\n\u003Cp>The new models are \u003Cstrong>available immediately\u003C\u002Fstrong> for developers and enterprises – via the Gemini API for custom integrations, Google Workspace for business users, and the Gemini app for individual users. This means organizations can deploy production-ready voice interfaces today without lengthy development cycles.\u003C\u002Fp>\n\u003Ch2>What this means for enterprises\u003C\u002Fh2>\n\u003Cp>For organizations globally, voice AI is becoming a commodity feature. Anyone planning customer support, internal assistants, or voice commerce solutions can now leverage proven, production-ready models without waiting for proprietary development. Gemini 3.8 Live Extended Thinking's benchmark leadership makes Google the preferred choice for raw voice quality. However, real-world performance on domain-specific tasks and industry-specific use cases remains to be tested in the coming weeks.\u003C\u002Fp>\n\u003Ch2>Sources\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fmodels-and-research\u002Fgemini-models\u002Fgemini-3-8-live-gemini-3-8-live-extended-thinking\u002F\">blog.google\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fdeepmind.google\u002Fblog\u002Fintroducing-gemini-3-8-live-and-3-8-live-extended-thinking\u002F\">Google DeepMind\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.unite.ai\u002Fgoogle-launches-gemini-3-8-live-and-extended-thinking-voice-models\u002F\">unite.ai\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fsiliconangle.com\u002F2026\u002F09\u002F15\u002Fgoogles-new-speech-model-gemini-3-8-live-supports-real-time-reasoning\u002F\">SiliconANGLE\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fthe-decoder.de\u002Fgoogle-deepmind-stellt-gemini-3-8-live-fuer-sprach-agenten-vor\u002F\">The Decoder (DE)\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cem>Editorially owned by \u003Ca href=\"\u002Fen\u002Fautor\u002Fideal-syka\">Ideal Syka\u003C\u002Fa>. Sources and method: \u003Ca href=\"\u002Fen\u002Fredaktion\">Newsroom & method\u003C\u002Fa>. Tips and corrections: \u003Ca href=\"mailto:ai@i6eal.de\">ai@i6eal.de\u003C\u002Fa>.\u003C\u002Fem>\u003C\u002Fp>\n",1789544829623]