Google launches Gemini 3.8 Live voice models at a fraction of GPT-Live-1's price

Google DeepMind released Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, which it calls its most advanced live dialogue models. Both are available through the Gemini API, and Google says they also reach Google Workspace and the Gemini app. 9to5Google reports the Extended Thinking variant now powers Gemini Live, Gmail, and Keep. The models handle real-time visual context and execute background tasks such as API calls without interrupting the conversation, and they support more than 97 languages. Google says the Extended Thinking variant ranks first on the Artificial Analysis Speech-to-Speech Leaderboard at 82.6 percent, ahead of OpenAI's GPT-Live-1 models. Pricing is the sharper contrast. Google charges $0.005 per minute for audio input and $0.018 for output, which The Decoder puts at about $1.38 for an hour of conversation, against at least $3.00 for GPT-Live-1 at $0.05 per minute. The Decoder notes that OpenAI's model is full duplex, listening and speaking at the same time, which may still give it more natural conversations, and judges from the demos that Google has again optimized for price over quality. Sample apps are on GitHub. For developers building voice agents, the release offers a much cheaper option less than a week after OpenAI opened its own voice model to the API.