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 |
| Muse Spark 1.3 | 58.1 |
| Muse Wonder Full | 58.1 |
| GPT-5.6 Sol | 53.5 |
| Muse Spark 1.1 | 51.2 |
| Grok 4.6 | 48.8 |
| Muse Wonder | 48.8 |
| Gemini 4 Argon | 46.5 |
| Gemini 4 Carbon | 46.5 |
| GLM-5.3 | 46.5 |
| Qwen3.8-Max | 46.5 |
| Qwen3.8 Max (0902) | 46.5 |
| Claude Haiku 5.5 | 44.2 |
| DeepSeek V4.1 Flash 1017 | 44.2 |
| GLM-5.3 Flash | 44.2 |
| Grok 4.7 | 44.2 |
| Hy4 Preview | 44.2 |
| MiMo-V2.6-Pro | 44.2 |
| Muse Spark 1.2 | 44.2 |
| Opus 4.8 | 41.9 |
| Gemini 3.8 Flash | 41.9 |
| Qwen3.8 27B | 41.9 |
| DeepSeek V4.1 Flash | 39.5 |
| GPT-6.1 Sol | 39.5 |
| Kimi K3 | 39.5 |
| Muse Spark Open | 39.5 |
| Step 5 Preview | 39.5 |
| Ember-1 | 37.2 |
| Gemini 3.7 Flash | 37.2 |
| GPT-5.5 | 37.2 |
| GPT-5.6 Terra | 37.2 |
| MiMo-V2.6-Flash | 37.2 |
| MiniMax M3.1 Flash | 37.2 |
| Sonnet 4.6 | 34.9 |
| DeepSeek V4 Pro 0813 | 34.9 |
| Inkling-Small | 34.9 |
| Opus 4.7 | 32.6 |
| Sonnet 5 | 32.6 |
| GPT-5.6 Luna | 32.6 |
| GPT-6 Astra | 32.6 |