Legal Research Bench Weighted Pass Immigration
Professional Work · 2026-06-23
Weighted-pass practice-area score for Immigration tasks in Vals AI Legal Research Bench.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Opus 5 | 86.9 |
| Muse Spark 1.1 | 82.5 |
| Grok 4.5 | 82.3 |
| GPT-5.5 | 82.1 |
| Kimi K3 | 82.0 |
| Inkling | 81.3 |
| GPT-5.6 Sol | 81.2 |
| Claude Fable 5 | 81.0 |
| GLM-5.1 | 80.7 |
| Gemini 3.5 Flash | 80.2 |
| GPT-5.6 Terra | 80.1 |
| Gemini 3.6 Flash | 80.1 |
| Sonnet 5 | 79.2 |
| MiniMax M3 | 78.6 |
| GPT-5.6 Luna | 78.6 |
| Opus 4.8 | 78.3 |
| GLM-5.2 | 78.1 |
| Sonnet 4.6 | 77.8 |
| Inkling-Small | 76.1 |
| DeepSeek-V4-Pro | 75.6 |
| Kimi K2.6 | 75.6 |
| Qwen3.7-Max | 73.3 |
| Gemini 3.1 Pro Preview | 72.9 |
| Gemini 3.5 Flash-Lite | 68.7 |
| GPT-5.4 Mini | 68.4 |
| Grok 4.3 | 67.7 |
| Laguna M.1 | 64.3 |
| Laguna XS.2 | 51.6 |