Skip to content

Benchmark

Vals LegalBench

Scores are percentages.

Models measured
133
Scores recorded
138
Sources
1
IQ weight
2.47

Score distribution

78.9%
Mean
82.4%
Median
9.2%
Std dev
68%
10th percentile
86.2%
90th percentile
33%
Min
33%88.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.47 for chat models, measured over 133 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Gemini 3.5 Flash· Vals AI83.6%
52Claude Opus 4.8· Vals AI83.6%
53Claude Opus 4.1· Vals AI83.5%
54GLM 4.7· Vals AI83.4%
55DeepSeek V4.1 Flash· Vals AI83.3%
56Grok 4· Vals AI83.2%
57Grok 3 Mini Fast Betahigh, reasoning· Vals AI83.1%
58GPT 4.1· Vals AI83.1%
59Claude Opus 4· Vals AI83.1%
60Mercury 2.5· Vals AI83.1%
61Inkling Small· Vals AI83%
62Claude Sonnet 4· Vals AI83%
63Inkling FP4· Vals AI82.9%
64GPT 5.2· Vals AI82.8%
65gemini-2.5-flash-preview-09-2025thinking· Vals AI82.6%
66Grok 3· Vals AI82.6%
67Claude Sonnet 3.7thinking· Vals AI82.5%
68gemini-2.5-flash-preview-09-2025· Vals AI82.5%
69Grok 4.1 Fast· Vals AI82.5%
70Qwen 3.8 27B· Vals AI82.4%
71DeepSeek V4 Pro 0813· Vals AI82.4%
72GPT 4o· Vals AI82.2%
73Claude Sonnet 4.6· Vals AI82.1%
74Nemotron 3 Ultra· Vals AI82.1%
75Claude Sonnet 4thinking· Vals AI82.1%
76Gemini 2.5 Flash Lite Preview 09-2025thinking· Vals AI82%
77Grok 3 Mini Fast Betalow, reasoning· Vals AI81.9%
78Qwen3 Max· Vals AI81.9%
79GPT-5 mini· Vals AI81.8%
80Kimi K2 Instruct· Vals AI81.5%
81Claude Haiku 4.5thinking· Vals AI81.2%
82DeepSeek V3· Vals AI80.8%
83Grok 4 Fast· Vals AI80.6%
84GPT-4 Turbo· Vals AI80.5%
85o1· Vals AI80.4%
86Qwen3 Max Preview· Vals AI80.3%
87DeepSeek V4 Pro· Vals AI80.3%
88Kimi K2 Thinking· Vals AI80.2%
89Qwen 3 235B A22B· Vals AI80.2%
90Claude Sonnet 3.7· Vals AI80%
91MiniMax M2.5· Vals AI80%
92Command A 03 2025· Vals AI79.7%
93GLM 4.6· Vals AI79.6%
94Qwen2.5 72B Instruct Turbo· Vals AI79.4%
95o4 Mini· Vals AI79.2%
96Mistral Large 3· Vals AI79.1%
97Gemini 2.5 Flash Lite Preview 09-2025· Vals AI79%
98MiMo V2.5· Vals AI78.9%
99Gemini 2.0 Flash· Vals AI78.4%
100GPT-4.1 mini· Vals AI78%

Showing 51–100 of 138 results.