Atlas

Benchmarks

← All benchmarks

Legal Research Bench Weighted Pass Civil Litigation

Professional Work · 2026-06-23

Weighted-pass practice-area score for Civil Litigation tasks in Vals AI Legal Research Bench.

Top models (higher is better)

ModelScore
Gemini 4 Carbon96.5
Gemini 4 Argon95.1
Muse Wonder Full92.9
Claude Opus 592.8
Gemini 3.8 Flash91.5
Muse Spark 1.391.2
Grok 4.691.1
Hy4 Preview90.8
Opus 4.890.8
MiniMax M3.1 Flash90.7
GPT-5.6 Sol90.2
Grok 4.790.0
Step 5 Preview90.0
GLM-5.389.9
GLM-5.3 Flash89.6
Muse Wonder89.0
Muse Spark 1.189.0
Gemini 3.7 Flash88.9
Grok 4.588.7
Sonnet 588.7
DeepSeek-V4-Flash-073188.7
GPT-6 Astra88.7
Muse Spark Open88.3
Opus 4.788.2
GPT-6.1 Sol87.7
MiMo-V2.6-Pro87.7
GPT-5.587.6
GPT-5.6 Terra87.5
Sonnet 4.687.4
Muse Spark 1.287.4
Ember-187.1
DeepSeek V4.1 Flash87.1
GLM-5.287.0
Gemini 3.5 Flash86.3
DeepSeek-V4-Pro86.0
Mistral Large 4 Preview85.8
DeepSeek V4 Pro 081385.8
Gemini 3.6 Flash85.3
Inkling85.0
MiniMax M384.9
Loading Atlas data…