BenchmarksCoding
KernelBench (Hard)
Stanford KernelBench's hardest tier: generate correct, high-performance GPU (CUDA) kernels from PyTorch reference operators, scored on the fraction of kernels that are both correct and faster than the baseline (fast₁).
Codingfast₁Higher is better
| Model | Lab | Score↓ | Source model | Status | Date | |
|---|---|---|---|---|---|---|
| MiniMax M3 | MiniMax | 28.8% | MiniMax M3 | Verified | — | Details |
Each row reports the model’s fast₁ on KernelBench (Hard). Click a row for the full run context.