Qwen 3.5 VLM just dropped — and it’s a very “agent-native” kind of multimodal

Beyond Words and Images - How Vision-Language-Action (VLA) Models Are Revolutionizing Robotics and Cyber-Physical Systems