media MarkTechPost · 13h ago · releases

Google releases Gemini 3.8 Live and Extended Thinking for production voice agents

from English

Google has introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, native speech-to-speech models designed for real-time voice agents that reason and execute tools without breaking conversational flow.

Both models are now live in the Gemini Live API and Google AI Studio as hosted solutions.
Gemini 3.8 Live focuses on scale and cost efficiency, while Extended Thinking adds multi-step reasoning capabilities.
Extended Thinking ranks #1 on Artificial Analysis’ Speech to Speech Quality Index with a score of 82.6.
The models support asynchronous function calling, visual context, alphanumeric precision, and 97 languages.
Pricing is set at $0.005/min for audio input and $0.018/min for audio output via the Live API.

These releases provide a streamlined alternative to cascaded speech pipelines, enabling developers to build production-grade voice agents with integrated tool use and low latency.

Importance 3/3 Beats a top-lab benchmark MarkTechPost Google DeepMind AI agents Voice & audio

Read original

arxiv arXiv cs.AI · 90d ago · 8 views

ScenA: Reference-Driven Multi-Speaker Audio Scene Generation

ScenA conditions a text-to-audio foundation model on multiple reference voices and a natural language scene prompt to generate realistic multi-speaker conversations. It addresses the 'Reference Shortcut' issue by using a high-noise-biased training schedule, ensuring speaker assignment relies on text prompts rather than acoustic similarity. Evaluated on CoVoMix2-Dialogue, Scen- A outperforms existing systems in speaker-binding and produces rich, naturalistic audio with overlapping speech and ambient noise.

blog Simon Willison · 12h ago · 1 view

Google releases Gemini 3.8 Live speech-to-speech models

Google has released two new speech-to-speech models, Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, which are similar in shape to OpenAI's GPT-Live models.

lab Google — The Keyword (AI) · 18h ago · 6 views

Google expands multilingual AI with Gemini 3.5, open datasets, and offline translation

Google reports its technologies now support over 300 languages for more than 7 billion people, introducing new tools to improve speech understanding, data collection for underrepresented languages, and accessibility.

lab Google — The Keyword (AI) · 18h ago · 6 views

Google reports AI advances in science, health, weather, and language access

Google announced that its technologies now support more than 300 languages spoken by 7 billion people and released new interactive insights via the AI & Economy ATLAS to track global AI usage. The company highlighted recent scientific breakthroughs, including the AlphaGenome Atlas mapping genetic changes and WeatherNext 3 delivering more accurate precipitation forecasts.

lab Google DeepMind Blog · 19h ago · 8 views

Google introduces Gemini 3.8 Live and 3.8 Live Extended Thinking

Google has released two new models, Gemini 3.8 Live and 3.8 Live Extended Thinking, designed to enhance near real-time reasoning for voice agents and improve the fluidity of AI conversations.