Qwen 3.7-Max Demonstrates Long-Running Autonomous Coding
Key Questions
What performance did Qwen 3.7-Max demonstrate in autonomous coding?
Qwen 3.7-Max ran autonomously for 35 hours and invoked over 1,000 tools to deliver a 10x faster kernel. It outperformed models like Opus 4.6, Gemini 3.1, and Deepseek v4 in testing.
What hardware accompanies the Qwen 3.7-Max release?
The release includes Alibaba's Zhenwu M890 chip and Panjiu server as supporting hardware components. These aim to enhance performance for long-running AI tasks.
How can sub-agents improve AI coding workflows?
Sub-agents allow splitting exploration from editing by handling research and codebase tasks in separate context windows. This keeps the primary session focused and improves efficiency.
Alibaba's Qwen 3.7-Max runs 35 hours autonomously, invokes 1,000+ tools to deliver 10x faster kernel. Accompanied by Zhenwu M890 chip and Panjiu server.
Sources (2)
Updated May 23, 2026