Google releases Gemini 3.8 Live audio models for developers
Google DeepMind has released Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, two audio models available to developers through the Gemini API and Google AI Studio. The models are designed for voice agents that can make API calls in the background, process visual input and continue speaking. Google says they support more than 97 languages, while sample applications are available on GitHub.
The Extended Thinking model leads the Artificial Analysis Speech-to-Speech Leaderboard with a score of 82.6 percent, ahead of OpenAI’s latest GPT-Live-1 models. Google charges $0.005 per minute for audio input and $0.018 per minute for output. That puts an hour of voice conversation at about $1.38, compared with at least $3.00 for GPT-Live-1.
The comparison also notes that OpenAI’s model may produce more natural conversations because its full-duplex system can listen and speak simultaneously. Demonstrations were described as sounding better, suggesting a trade-off between Google’s lower price and conversational quality.
Written by Sakhanda Wire AI from reporting by The Decoder. Facts, quotes and full context are in the original article below.