Horangi 4 - GLP - Logical Reasoning: leaderboard
Logical reasoning category of the Horangi Korean LLM Leaderboard 4 (W&B Korea): the mean of Ko-HellaSwag accuracy and Ko-MT-Bench's reasoning-category judge rating (divided by 10). 0-100, higher is better.
Metric: Score (%). Source: horangi.ai. Saturation forecast: Estimated already saturated. 105 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) (High) | 90.5 | #64 (Gemini 3.1 Pro (Preview)) |
| 2 | Qwen 3.7 Plus | 90.11 | #98 |
| 3 | Claude Opus 4.1 (20250805) | 89.5 | #139 |
| 4 | Claude Sonnet 4.5 (High) | 89.5 | #148 (Claude Sonnet 4.5) |
| 5 | Claude Opus 4.5 (20251101) (High) | 89.5 | #81 (Claude Opus 4.5 (20251101)) |
| 6 | Qwen 3.7 Max | 89.17 | #85 |
| 7 | Claude Opus 4.6 (High) | 89 | #67 (Claude Opus 4.6) |
| 8 | GPT-5.1 (2025-11-13) | 88.5 | #107 |
| 9 | Qwen 3.5 27B | 87.5 | #192 |
| 10 | Gemini 3 Flash (Preview) | 87.5 | #90 |
| 11 | Gemini 3.5 Flash | 87.5 | #66 |
| 12 | GPT-5.4 (2026-03-05) (xHigh) | 87.5 | #63 (GPT-5.4 (2026-03-05)) |
| 13 | GLM-5.2 (High) | 87.5 | #80 (GLM-5.2) |
| 14 | Kimi K2.5 | 87 | #152 |
| 15 | Claude Opus 4.8 (High) | 87 | #39 (Claude Opus 4.8) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=horangi-4-glp-logical-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-05.