IndustryBench (Thinking): leaderboard
Metric: Final (SV) score (out of 3): mean 0-3 rubric score from a Qwen3-Max judge against the reference answer, set to 0 for any answer with a safety violation against the source standard; closed-book, zero-shot, thinking mode enabled, only the final answer judged, over the 2,049 Chinese industrial procurement QA items of IndustryBench grounded in Chinese national standards (GB/T); higher is better. Source: arxiv.org. Saturation forecast: Around May 2027. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 (Thinking) | 2.03 |
| 2 | GPT-5.4 (Thinking) | 1.98 |
| 3 | Gemini 3.1 Pro (Preview) | 1.97 |
| 4 | Qwen 3.6 Plus (Thinking) | 1.89 |
| 5 | Qwen 3.5 397B A17B | 1.81 |
| 6 | Qwen 3.5 Plus (Thinking) | 1.79 |
| 7 | Qwen 3 Max (Thinking) | 1.75 |
| 8 | GLM-5 | 1.72 |
| 9 | Qwen 3.5 122B A10B | 1.71 |
| 10 | Kimi K2.5 | 1.68 |
| 11 | Qwen 3.5 27B | 1.65 |
| 12 | Qwen 3.5 35B A3B | 1.64 |
| 13 | MiniMax-M2.5 | 1.42 |
Interactive version: theaggregate.ai/benchmark?slug=industrybench-thinking · How It Works · Data refreshed daily, snapshot 2026-10-07.