Legal Research Bench Business/Commercial
Professional Work · 2026-06-23
This row reports the practice-area score for Business/Commercial tasks in Vals AI Legal Research Bench.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Opus 5 | 62.8 |
| GPT-5.6 Sol | 53.5 |
| Claude Fable 5 | 51.2 |
| Muse Spark 1.1 | 51.2 |
| Kimi K3 | 46.5 |
| Opus 4.8 | 41.9 |
| GPT-5.5 | 37.2 |
| GPT-5.6 Terra | 37.2 |
| Sonnet 4.6 | 34.9 |
| Inkling-Small | 34.9 |
| Sonnet 5 | 32.6 |
| GPT-5.6 Luna | 32.6 |
| Grok 4.5 | 30.2 |
| Gemini 3.5 Flash | 27.9 |
| GLM-5.2 | 27.9 |
| Inkling | 25.6 |
| MiniMax M3 | 25.6 |
| DeepSeek-V4-Pro | 23.3 |
| Gemini 3.1 Pro Preview | 23.3 |
| Gemini 3.6 Flash | 23.3 |
| GLM-5.1 | 23.3 |
| Kimi K2.6 | 20.9 |
| Qwen3.7-Max | 20.9 |
| Grok 4.3 | 14.0 |
| GPT-5.4 Mini | 11.6 |
| Gemini 3.5 Flash-Lite | 7.0 |
| Laguna M.1 | 4.7 |
| Laguna XS.2 | 0.0 |