LLM2014 Logic 2025-05 — leaderboard
Metric: Median Score. Source: raw.githubusercontent.com. 32 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O3 (High) | 85.14 |
| 2 | O4 Mini (High) | 80.16 |
| 3 | Grok 3 Mini | 69.06 |
| 4 | Gemini 2.5 Pro (Preview 05-06) | 68.33 |
| 5 | Claude Sonnet 4 (Thinking) | 68.22 |
| 6 | Qwen 3 235B A22B (Thinking) | 66.6 |
| 7 | Claude Opus 4 | 60.12 |
| 8 | GPT-4.1 Mini | 53.52 |
| 9 | Claude Sonnet 4 | 52.08 |
| 10 | Qwen 3 30B A3B (Thinking) | 51.6 |
| 11 | DeepSeek V3 (0324) | 46.2 |
| 12 | Gemini 2.5 Flash | 42.85 |
| 13 | Qwen 3 235B A22B | 40.18 |
| 14 | Gemini 2.5 Flash (Preview 05-20) | 39.06 |
| 15 | GPT-4.1 | 36.62 |
Interactive version: theaggregate.ai/benchmark?slug=llm2014-logic-2025-05 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.