Google-led researchers introduce StreamVibe proactive video assistant
A Google-led research team introduced StreamVibe, a video agent that observes activities and decides when to speak, including for vague requests to provide help when difficulties arise. The team trained StreamVibe-VL, apparently based on Alibaba's Qwen3.5-9B, alongside reinforcement learning and a memory system for tracking requests and completed steps. On a benchmark of about 4,100 examples drawn from more than 183 hours of recordings, the researchers reported 10.7% better response timing and quality and a 12.3% boost in proactive interaction from memory, though Gemini 3.8 Flash scored higher in several tests and performance deteriorated as continuous videos grew longer.
The research targets a practical limitation in video assistants: determining when intervention is useful while tracking explicit, implicit, and changing user intentions over time.
Google wants video AI to understand what people need as they work
Jonathan Kemper View the LinkedIn Profile of Jonathan Kemper
Nov 1, 2026

Liu et al.
A research team led by Google has introduced StreamVibe, a video agent designed to decide for itself when to speak up. The tests show above all how hard that remains for today's AI models.
#Google#Stanford University#University of Oxford#StreamVibe#StreamVibe-VL#Qwen3.5-9B#Gemini 2.5 Flash#Gemini 3.8 Flash