Alibaba's Qwen3.7-Plus delivers 1M context and agent-grade intelligence

Alibaba has launched Qwen3.7-Plus, the latest addition to its Qwen3.7 model series, now accessible through the Bailian API platform. This model supports multimodal inputs—both text and images—and produces text outputs, representing a significant upgrade to its vision-language capabilities. With a 1 million token context window, Qwen3.7-Plus is positioned as a full-stack, agent-level AI for coding, tool use, and productivity workflows. Benchmarks indicate strong performance in overall intelligence, coding benchmarks, and agentic tasks, making it a competitive option in the enterprise AI space.

A notable demonstration of Qwen3.7-Plus's autonomous capabilities involved the model independently generating an English vocabulary learning application over 11 continuous hours, showcasing its ability to handle extended, complex projects without human intervention. By offering a cost-effective API, Alibaba targets developers and businesses looking to integrate advanced AI agents for prototyping, automation, and educational tools, challenging existing leaders like OpenAI and Anthropic in the agentic AI market.