Alibaba's Qwen3.7-Max AI runs autonomously for 35 hours on unknown hardware, makes 1,158 tool calls, achieves 10x speedup on kernel optimization—outperforming DeepSeek, GLM, and Kimi through environment-scaled training.

Alibaba's Qwen3.7-Max Ran Autonomously for 35 Hours on Unfamiliar Hardware. It Still Kept Getting Better. - Firethering

- Advertisement - Alibaba gave Qwen3.7-Max a kernel optimization task on a hardware platform the model had never encountered before. No documentation or profiling data. No example kernels for the architecture. Just a task description, an existing implementation, and an evaluation script. The model ran for 35 hours. It made 1,158 tool calls. It wrote, compiled, profiled, and rewrote the kernel repeatedly, diagnosing failures, fixing bugs, identifying blocks, and redesigning the architecture multi...