LLM2014 Logic 2025-06: leaderboard
LLM2014 monthly logic benchmark for 2025-06, reporting aggregate logical-reasoning performance on the source score scale.
Metric: Median Score. Source: raw.githubusercontent.com. Saturation forecast: Estimated already saturated. 38 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O3 (High) | 83.95 | #124 (O3) |
| 2 | O4 Mini (High) | 78.97 | #180 (O4 Mini) |
| 3 | Seed-1.6 (Thinking) | 71.7 | #265 (Seed-1.6) |
| 4 | Gemini 2.5 Pro | 71.36 | #152 |
| 5 | DeepSeek R1 0528 | 66.99 | #222 |
| 6 | Qwen 3 235B A22B (Thinking) | 63.05 | #313 (Qwen 3 235B A22B) |
| 7 | Claude Sonnet 4 (Thinking) | 62.96 | #199 (Claude Sonnet 4) |
| 8 | Grok 3 Mini | 60.72 | #281 |
| 9 | Gemini 2.5 Flash (Thinking) | 55.46 | #245 (Gemini 2.5 Flash) |
| 10 | Claude Opus 4 | 53.48 | #163 |
| 11 | Qwen 3 32B (Thinking) | 50.72 | #443 (Qwen 3 32B) |
| 12 | GPT-4.1 Mini | 48.53 | #355 |
| 13 | SenseNova V6 Reasoner | 47.91 | #306 |
| 14 | Qwen 3 30B A3B (Thinking) | 47.41 | #493 (Qwen 3 30B A3B) |
| 15 | Claude Sonnet 4 | 45.97 | #199 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm2014-logic-2025-06 · How It Works · Data refreshed daily, snapshot 2026-10-20.