CaseLaw v2: leaderboard
Private question-answer benchmark over Canadian court-cases.
Metric: Accuracy (%). Source: vals.ai. Source: benchmarklist.com. 54 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Grok 4.3 | 79.31 | #188 |
| 2 | GPT-5.1 (2025-11-13) (High) | 73.42 | #101 (GPT-5.1 (2025-11-13)) |
| 3 | GPT-4.1 | 69.88 | #248 |
| 4 | GPT-5 Mini (2025-08-07) (High) | 68.49 | #172 (GPT-5 Mini (2025-08-07)) |
| 5 | Claude Opus 4.7 | 68.38 | #49 |
| 6 | GPT-5 | 66.45 | #98 |
| 7 | GPT-5 (2025-08-07) (High) | 66.45 | |
| 8 | GPT-5.5 (xHigh) | 66.24 | #31 (GPT-5.5) |
| 9 | GPT-5.2 | 66.02 | #110 |
| 10 | GPT-5.2 (2025-12-11) (xHigh) | 66.02 | |
| 11 | Grok 4 | 65.81 | #181 |
| 12 | Kimi K2 (Thinking) | 65.7 | #243 (Kimi K2) |
| 13 | Grok 4 Fast (Reasoning) | 65.7 | #254 (Grok 4 Fast) |
| 14 | Gemini 3.1 Pro (Preview) (High) | 64.84 | #58 (Gemini 3.1 Pro (Preview)) |
| 15 | Command A | 64.52 | #508 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=caselaw-v2 · How It Works · Data refreshed daily, snapshot 2026-10-20.