PrinzBench: leaderboard
Private benchmark testing LLMs on legal research, analysis, and locating obscure public information. 33 questions (25 legal, 8 search) scored out of 99.
Metric: Score (x/99). Source: github.com. Status: saturation imminent. 29 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.6 Pro Sol | 91 |
| 2 | GPT-5.4 (xHigh) | 69 |
| 3 | GPT-5.3 Codex (High) | 52 |
| 4 | Gemini 3.1 Pro (Preview) | 50 |
| 5 | Kimi K3 | 47 |
| 6 | Grok 4.20 | 43 |
| 7 | Gemini 3 Flash | 36 |
| 8 | Gemini 3 Pro | 35 |
| 9 | Kimi K2.5 (Thinking) | 35 |
| 10 | Muse Spark | 31 |
| 11 | GLM-5.2 | 30 |
| 12 | Claude Opus 4.7 | 25 |
| 13 | Qwen 3 Max | 25 |
| 14 | Grok 4 | 23 |
| 15 | DeepSeek V4 Pro | 23 |
Interactive version: theaggregate.ai/benchmark?slug=prinzbench · How It Works · Data refreshed daily, snapshot 2026-09-05.