PTXBench (H100) - GEMM: leaderboard

Metric: GEMM on H100: target-instruction turn correctness rate (%) within eight turns: share of MiniPTXAgent turns yielding a functionally correct CUDA kernel with inline PTX that runs a selected architecture-specific instruction at runtime (NCU-verified); three prompt variants times four eight-turn trajectories (96 turns), architecture knowledge pack provided. Source: arxiv.org. Saturation forecast: Around December 2026. 4 models tracked.

Top models

#ModelScore
1Claude Opus 4.8 (xHigh)94.8
2GLM-5.261.5
3Gemini 3.1 Pro (Preview)56.2
4Qwen 3.6 27B0

Interactive version: theaggregate.ai/benchmark?slug=ptxbench-h100-gemm · How It Works · Data refreshed daily, snapshot 2026-09-29.