SuperCLUE General (March 2025) - Text Understanding and Creation: leaderboard

Metric: Score. Source: www.superclueai.com. 36 models tracked.

Top models

#ModelScore
1QwQ-32B82.27
2Gemini 2.0 Flash (01-21) (Thinking)81.16
3DeepSeek R180.41
4GPT-4.5 (Preview)80.4
5Step 2 16K80.17
6Gemini 2.0 Pro (Preview 02-05)79.34
7GLM-4 Plus79.17
8O3 Mini (High)79.01
9DeepSeek V378.99
10hunyuan-turbos-2025022678.88
11Claude 3.7 Sonnet78.77
12Qwen 2.5 72B Instruct78.52
13Doubao-1.5-pro-32k-25011577.66
14DeepSeek R1 Distill Qwen 32B77.53
15Qwen 2.5 14B Instruct76.87

Interactive version: theaggregate.ai/benchmark?slug=superclue-general-march-2025-text-understanding-and-creation · How It Works · Data refreshed daily, snapshot 2026-09-19.