Alibaba's Qwen3.7-Max Runs Autonomously for 35 Hours

Alibaba has unveiled its proprietary Qwen3.7-Max model, demonstrating a remarkable capability to run autonomously for 35 continuous hours. This achievement marks a significant milestone in what the industry is calling the "agent era" of artificial intelligence, where models move beyond simple question-answering to actively plan, execute, and course-correct complex tasks over extended periods. The Qwen3.7-Max model is designed to operate with minimal human intervention, making decisions and adjusting its approach as circumstances change. During the 35-hour demonstration, the model reportedly managed complex workflows, handled unexpected errors, and dynamically reprioritized tasks without requiring human input. This level of autonomy represents a substantial leap forward from earlier AI models that typically required frequent human oversight for anything beyond short, well-defined tasks. A particularly notable feature of Qwen3.7-Max is its support for external "harnesses" such as Anthropic's Claude Code. This interoperability allows the model to leverage specialized tools and frameworks developed by other companies, potentially creating a more flexible and powerful AI ecosystem. By supporting external harnesses, Alibaba is signaling a commitment to openness and collaboration, even as it develops proprietary technology. The implications of extended autonomous operation are far-reaching. For businesses, AI agents that can work independently for days could automate complex processes in fields like data analysis, software development, supply chain management, and scientific research. Instead of requiring constant human supervision, these agents could be assigned long-term objectives and trusted to work through the details. However, the development also raises important questions about control, safety, and accountability. As AI systems become more autonomous, ensuring they remain aligned with human intentions and can be reliably stopped or redirected becomes increasingly critical. Alibaba's achievement demonstrates the technical feasibility of long-duration autonomous AI, but it also underscores the need for robust governance frameworks to manage such powerful capabilities.