OpenCompass Math - College - Chinese (CompassBench 2510): leaderboard

OpenCompass CompassBench 2510 edition (file dated 2025-11-10): college mathematics sub-task of the math category, Chinese questions. CompassBench is OpenCompass's periodic evaluation of chat models on a private question set; each edition is a separate board. Score 0-100, higher is better.

Metric: Score (%). Source: rank.opencompass.org.cn. 27 models tracked.

Top models

#ModelScoreOverall rank
1Gemma 3 27B (IT)56#528
2Grok 456#189
3GLM-4.656#265
4MiniMax-M256#327
5DeepSeek V3.2 Exp56#241
6Ling-1T56#354
7Hunyuan-T1-2025082256#306
8Qwen 3 Next 80B A3B (Thinking)52#312
9Claude Sonnet 4.5 (Thinking)52#148 (Claude Sonnet 4.5)
10DeepSeek R1 052852#232
11Kimi K2 090552#292
12Llama 4 Maverick Instruct52#467
13Claude Haiku 4.5 (Thinking)52#287 (Claude Haiku 4.5)
14O4 Mini (2025-04-16) (High)52#183 (O4 Mini (2025-04-16))
15Qwen 3 235B A22B 2507 (Thinking)48#263

Interactive version: theaggregate.ai/benchmark?slug=opencompass-math-college-chinese-compassbench-2510 · How It Works · Data refreshed daily, snapshot 2026-11-05.