China’s MiniMax And DeepSeek Have A Compute Capacity Of Sub-100,000 H100-Equvalent GPUs, And Yet Their Latest Models Continue To Scale The Frontier Curve, With Distillation Failing To Explain This Disparity

The heretofore uncontested placement of the DeepSeek-V4.1-Flash-1017 model at the pareto on the agentic benchmark score vs. cost curve (which simply means that the model achieves an optimal balance between the two) was fairly short-lived, with the new MiniMax…

Preview courtesy of Wccftech. The full article opens on their site.