LexRubric - Reasoning and Logic: leaderboard
Metric: Dimension score rate (%): points earned from expert-written atomic rubric items of the reasoning and logic dimension (weighted -10 to +10; negative items subtract) divided by the maximum positive score, averaged over instances; a Qwen3.6-27B judge checks each item, temperature 0.6, 16k output tokens; can fall below 0; higher is better (instances without a positive item in the dimension are left out). Source: arxiv.org. Saturation forecast: Around September 2027. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Kimi K2.6 | 79.08 |
| 2 | Qwen 3.6 Max Preview | 77.14 |
| 3 | Qwen 3 Max | 77.07 |
| 4 | Kimi K2.5 | 76.5 |
| 5 | GLM-5.1 | 74.72 |
| 6 | Claude Sonnet 4.6 | 72.24 |
| 7 | Qwen 3.5 397B A17B | 72.17 |
| 8 | GLM-5 | 68.24 |
| 9 | DeepSeek V4 Pro | 65.86 |
| 10 | DeepSeek V4 Flash | 65.71 |
| 11 | GPT-5.2 | 63.87 |
| 12 | DeepSeek R1 | 62.43 |
| 13 | DeepSeek V3.2 | 61.36 |
Interactive version: theaggregate.ai/benchmark?slug=lexrubric-reasoning-and-logic · How It Works · Data refreshed daily, snapshot 2026-09-29.