Qwen-3.6-27B and Gemma-4-31B Surpass Claude Mythos with Scaled Compute

Core summary

One sentence to understand this update

A user claims that by scaling test-time compute for Qwen-3.6-27B and Gemma-4-31B, they surpassed Claude Mythos in code optimizations and speedups, albeit utilizing approximately 25-40x more compute.

Impact & opportunity

What this could mean

This highlights that strategic compute scaling can enable smaller models to outperform top-tier closed-source models in specific tasks, encouraging builders to explore compute optimization strategies.