IndustryBench - Standards and Terminology: leaderboard

IndustryBench (2026) restricted to its 610 Standards and Terminology questions (national standard citation, industry terms and technical specifications), Chinese industrial procurement QA answered closed-book in default non-reasoning mode and scored 0 to 3 by a Qwen3-Max judge, with answers that violate a safety clause set to 0. Mean score out of 3; higher is better.

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, provider-default decoding, over the 610 items of the 2,049 Chinese industrial procurement QA items of IndustryBench grounded in Chinese national standards (GB/T) in the Standards and Terminology capability dimension (national standard citation, industry terms and technical specifications; default non-reasoning runs); higher is better. Source: arxiv.org. Saturation forecast: Around May 2027. 17 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3.1 Pro (Preview)1.76#99
2Qwen 3.6 PlusNon-reasoning1.65#238
3Claude Opus 4.61.62#113
4Kimi K2.5Non-reasoning1.61#266
5GPT-5.4Non-reasoning1.57#135
6Qwen 3.5 PlusNon-reasoning1.55#242
7Qwen 3.5 397B A17BNon-reasoning1.55#264
8Qwen 3.5 122B A10BNon-reasoning1.52#346
9GLM-5Non-reasoning1.5#269
10Qwen 3 MaxNon-reasoning1.5#415
11GPT-5.2Non-reasoning1.48#185
12Claude Sonnet 4.61.41#173
13MiniMax-M2.51.39#600
14Qwen 3.5 27BNon-reasoning1.36#375
15Qwen 3.5 35B A3BNon-reasoning1.21#492

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Example tasks

在国家标准GB/T 4324.2—2023中,所规定测定的对象包括哪些元素?

Source

自动换刀7:24圆锥工具柄的锥角公差和形状误差应依据哪些标准文件进行规定?

Source

Interactive version: theaggregate.ai/benchmark?slug=industrybench-standards-and-terminology · How It Works · Data refreshed daily, snapshot 2026-11-21.