SuperCLUE General (September 2025) - Hallucination Control: leaderboard

Metric: Score. Source: www.superclueai.com. 37 models tracked.

Top models

#ModelScore
1Claude Opus 4.1 (Thinking)85.23
2Gemini 3 Pro (Preview)81.51
3Claude Sonnet 4.5 (Thinking)78.61
4GPT-5 (High)78.34
5Hunyuan-T1-2025082277.96
6Doubao-Seed-1.6-thinking-25071577.78
7O3 (High)77.47
8GPT-5.1 (High)77.11
9Grok 473.52
10Gemini 2.5 Pro73.17
11GLM-4.673.08
12Gemini 2.5 Flash72.36
13Kimi K2 (Thinking)72.23
14DeepSeek V3.1 Terminus71.44
15DeepSeek V3.2 Exp (Thinking)71.26

Interactive version: theaggregate.ai/benchmark?slug=superclue-general-september-2025-hallucination-control · How It Works · Data refreshed daily, snapshot 2026-09-19.