SuperCLUE-Fact - Overall: leaderboard
SuperCLUE-Fact evaluates factuality hallucination in Chinese LLMs. Headline board of the newest workbook the superclueai.com vertical lists; score 0-100, higher is better.
Metric: Score. Source: www.superclueai.com. Saturation forecast: Estimated already saturated. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek R1 | 86.02 | #262 |
| 2 | GPT-4.5 (Preview) | 85.3 | #260 |
| 3 | Gemini 2.5 Pro (Preview 03-25) | 84.95 | #126 |
| 4 | DeepSeek V3 (0324) | 82.8 | #349 |
| 5 | Doubao-1.5-Pro-32k | 81.72 | #310 |
| 6 | QwQ-32B | 78.85 | #422 |
| 7 | O3 Mini (High) | 78.78 | #282 (O3 Mini) |
| 8 | Gemini 2.0 Flash | 78.49 | #351 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=superclue-fact-overall · How It Works · Data refreshed daily, snapshot 2026-11-05.