Atlas

Benchmarks

← All benchmarks

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)

ModelScore
Claude Opus 562.8
GPT-5.6 Sol53.5
Claude Fable 551.2
Muse Spark 1.151.2
Kimi K346.5
Opus 4.841.9
GPT-5.537.2
GPT-5.6 Terra37.2
Sonnet 4.634.9
Inkling-Small34.9
Sonnet 532.6
GPT-5.6 Luna32.6
Grok 4.530.2
Gemini 3.5 Flash27.9
GLM-5.227.9
Inkling25.6
MiniMax M325.6
DeepSeek-V4-Pro23.3
Gemini 3.1 Pro Preview23.3
Gemini 3.6 Flash23.3
GLM-5.123.3
Kimi K2.620.9
Qwen3.7-Max20.9
Grok 4.314.0
GPT-5.4 Mini11.6
Gemini 3.5 Flash-Lite7.0
Laguna M.14.7
Laguna XS.20.0
Loading Atlas data…