TRLawBench — leaderboard
Turkish legal benchmark testing legal knowledge, reasoning, and open-ended analysis over Turkish law tasks.
Metric: Stage 1 accuracy (self-reported). Source: benchmarklist.com. Status: saturation imminent. 32 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash | 85 |
| 2 | O4 Mini | 84.4 |
| 3 | Gemini 2.5 Pro | 83.8 |
| 4 | O3 | 81.3 |
| 5 | GPT-4.1 | 80 |
| 6 | Claude 3.7 Sonnet | 79.4 |
| 7 | GPT-4.1 Mini | 76.3 |
| 8 | Claude Sonnet 4 | 75.6 |
| 9 | Claude 3.7 Sonnet (Thinking) | 75.6 |
| 10 | GPT-4o | 75 |
| 11 | Claude 3.5 Sonnet | 74.4 |
| 12 | Qwen 3 235B A22B | 72.5 |
| 13 | Mistral Large | 71.3 |
| 14 | Gemma 3 27B | 70 |
| 15 | Llama 3.3 70B Instruct | 69.4 |
Interactive version: theaggregate.ai/benchmark?slug=trlawbench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.