LiveKit has made Google's open-weight Gemma 4 31B model available on its Inference platform in a version tuned specifically for real-time voice agents, pairing sub-400-millisecond response times with tool-use scores that the companies say edge out GPT-4.1.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.