IndustryCode - Python Sub-Problems: leaderboard
Metric: Pass@1 (%) on the Python sub-problems of the IndustryCode test split (106 main problems decomposed into 499 sub-problems; industrial Python, C++, MATLAB and Stata projects with test cases), each sub-problem generated with the main-problem description and the code of earlier steps as context; higher is better. Source: arxiv.org. Saturation forecast: Around April 2027. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.5 | 61.8 |
| 2 | Gemini 3 Pro | 59.7 |
| 3 | Claude Sonnet 4.5 | 58.6 |
| 4 | Qwen 3 Max | 56.6 |
| 5 | Gemini 2.5 Pro | 54.5 |
| 6 | Claude Haiku 4.5 | 53.4 |
| 7 | Kimi K2 (Thinking) | 51.8 |
| 8 | GLM-4.7 (Thinking) | 50.3 |
| 9 | Qwen 3 Coder Plus | 50 |
| 10 | DeepSeek V3.2 Exp | 49.5 |
| 11 | Kimi K2 | 48.7 |
| 12 | GPT-5.2 | 48.5 |
| 13 | GPT-5.1 | 48.5 |
| 14 | Doubao-Seed-1.6 (Thinking) | 44.5 |
| 15 | Grok 4.1 | 42.4 |
Interactive version: theaggregate.ai/benchmark?slug=industrycode-python-sub-problems · How It Works · Data refreshed daily, snapshot 2026-10-07.