LingOly-TOO: leaderboard
Linguistics reasoning benchmark evaluating models on baseline and obfuscated questions to separate reasoning ability from memorization.
Metric: Obfuscated Score (self-reported). Source: benchmarklist.com. Status: saturation imminent. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 46.7 |
| 2 | Claude Opus 4.1 | 45.8 |
| 3 | Claude 3.7 Sonnet | 42.89 |
| 4 | Gemini 2.5 Pro | 42.35 |
| 5 | DeepSeek V3.1 Terminus | 42.2 |
| 6 | O1 Preview | 32.22 |
| 7 | O3 Mini (High) | 30.59 |
| 8 | Claude 3.5 Sonnet | 28.1 |
| 9 | DeepSeek R1 | 26.5 |
| 10 | GPT-4.5 | 25.45 |
| 11 | Gemini 1.5 Pro | 20.46 |
| 12 | GPT-4o | 15.63 |
| 13 | O3 Mini (Low) | 12.22 |
| 14 | Phi-4 | 11 |
| 15 | Llama 3.3 70B Instruct | 8.21 |
Interactive version: theaggregate.ai/benchmark?slug=lingoly-too · How It Works · Data refreshed daily, snapshot 2026-09-05.