LLMEval-Logic Hard Sub-Q — leaderboard
Metric: Accuracy (%). Source: llmeval.com. 14 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 (Thinking) | 76.6 |
| 2 | Gemini 3.1 Pro (Preview) (Thinking) | 76 |
| 3 | Claude Opus 4.6 (Non-reasoning) | 74.4 |
| 4 | Gemini 3.1 Pro (Preview) | 71.3 |
| 5 | GPT-5.4 Pro (xHigh) | 70 |
| 6 | Qwen 3.5 Plus (Thinking) | 70 |
| 7 | Kimi K2.5 (Thinking) | 68.4 |
| 8 | Hy3-preview (Reasoning) | 62.7 |
| 9 | Hy3-preview (Non-reasoning) | 27.4 |
| 10 | Kimi K2.5 (Non-reasoning) | 27 |
Interactive version: theaggregate.ai/benchmark?slug=llmeval-logic-hard-sub-q · How the rankings work · Data refreshed daily, snapshot 2026-07-22.