LLM2014 Logic 2026-03: leaderboard
LLM2014 monthly logic benchmark for 2026-03, reporting aggregate logical-reasoning performance on the source score scale.
Metric: Median Score. Source: raw.githubusercontent.com. Saturation forecast: Around January 2027. 42 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 (High) | 78.85 | #79 (GPT-5.4) |
| 2 | Claude Opus 4.6 (Thinking) | 77.23 | #63 (Claude Opus 4.6) |
| 3 | Gemini 3.1 Pro (Preview) (High) | 75.02 | #58 (Gemini 3.1 Pro (Preview)) |
| 4 | Seed 2.0 Pro (High) | 68.36 | #118 (Seed 2.0 Pro) |
| 5 | Gemini 3 Flash (High) | 65.29 | #99 (Gemini 3 Flash) |
| 6 | Kimi K2.5 (Thinking) | 59.56 | #144 (Kimi K2.5) |
| 7 | GLM-5 (Thinking) | 55.08 | #141 (GLM-5) |
| 8 | GPT-5.4 Mini (High) | 54.45 | #212 (GPT-5.4 Mini) |
| 9 | Claude Opus 4.6 | 53.93 | #63 |
| 10 | Claude Sonnet 4.5 (Thinking) | 52.81 | #142 (Claude Sonnet 4.5) |
| 11 | Qwen 3.5 Plus (Thinking) | 51.81 | #125 (Qwen 3.5 Plus) |
| 12 | MiniMax-M2.7 | 49.52 | #258 |
| 13 | Qwen 3.5 27B (Thinking) | 47.86 | #187 (Qwen 3.5 27B) |
| 14 | Step 3.5 Flash | 39.99 | #270 |
| 15 | Qwen 3.5 Plus | 39.29 | #125 |
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-2026-03 · How It Works · Data refreshed daily, snapshot 2026-10-20.