Google Challenges OpenAI with Affordable Gemini 3.8 Live Audio Models
Google Deepmind has released Gemini 3.8 Live and its Extended Thinking variant, two new audio models for developers through the Gemini API and Google AI Studio. These models allow voice agents to make API calls in the background while processing visual input and keeping a conversation going.
Gemini 3.8 Live supports over 97 languages, and its Extended Thinking variant ranks first on the Artificial Analysis Speech-to-Speech Leaderboard with an impressive 82.6% score, outperforming OpenAI's latest GPT-Live-1 models.
What sets these models apart is their affordability: Google charges $0.005 per minute for audio input and $0.018 for output, significantly cheaper than OpenAI's GPT-Live-1 at $0.05 per minute. An hour-long voice conversation with Gemini 3.8 Live would cost approximately $1.38, while the same conversation with OpenAI's model would set you back at least $3.00.
While Google's models may not offer as natural-sounding conversations as OpenAI's due to their lack of full duplex capabilities, they still provide a viable alternative for developers looking for cost-effective solutions.