Alibaba releases Qwen3.7-Plus, a multimodal AI agent that excels at operating graphical interfaces and autonomous app development.

Alibaba's Qwen team released Qwen3.7-Plus, a multimodal model combining visual perception with agent capabilities like coding and tool use. In demos, the model ran for over eleven hours to build an English vocabulary learning app, producing more than 10,000 lines of code across more than 1,000 agent calls; it also autonomously recreated a macOS Stocks app and handled cloud console tasks like purchasing virtual server instances.

Qwen3.7-Plus excels at operating graphical interfaces and leads on AndroidWorld and ScreenSpot Pro benchmarks, outperforming GPT-5.4 (xhigh), Opus 4.6 Max, and Gemini 3.1 Pro. However, on pure reasoning tasks like MedXpertQA-MM, it falls behind Gemini 3.1 Pro and GPT-5.4. The model supports Anthropic's API protocol and works with Claude Code, OpenClaw, and Qwen Code.

Qwen3.7-Plus is priced at $0.40 per million input tokens and $2.40 per million output tokens, compared to $2.50 and $7.50 for Qwen3.7-Max. The model is available through Alibaba Cloud Model Studio and is proprietary with no open weights.