Alibaba Launches Qwen3.7-Plus: Multimodal Agent Model With Vision, Deep Reasoning, and Tool Invocation

Qwen / Hacker News·medium signal

Alibaba's Qwen team released Qwen3.7-Plus, adding image/video understanding, deep reasoning, self-programming, tool invocation, verification, and autonomous iteration to the Qwen3.7 base. The model works across Claude Code, OpenClaw, and Qwen Code frameworks with cross-framework generalization. Ranked #16 overall in Vision Arena, making Alibaba the #5 lab in vision. Available now on Alibaba Cloud Model Studio. 40 points on HN.