Google introduces StreamVibe, a video agent that decides when to speak during tasks

A research team led by Google has introduced StreamVibe, a video agent designed to decide for itself when to speak up during tasks such as cooking or assembling furniture. The model, based on Alibaba's Qwen3.5-9B, evaluates every second whether to stay quiet or offer help, and uses a separate memory system to track pending requests and completed steps. Tests show StreamVibe-VL improves response timing and quality by 10.7 percent over other models, while the memory system boosts proactive interaction by another 12.3 percent and reduces unnecessary interruptions.