HLAB-Vals 1.1 (legal agent)

Legal work done by an agent with documents, spreadsheets, presentations and file tools. Run by Vals AI.

CategoryAgents
MetricAccuracy (%)
SourceVals AI (evaluator)
LicenseNone stated
Benchmark pagewww.vals.ai
StatusScored
Weight0.8 within Agents
FitDiscrimination 0.59, noise 0.69

Results

#ModelProviderResultDate
1 Gemini 4 Argon High Google 27.1% 2026-11-09
2 Muse Spark Open Extra High Meta 25.0% 2026-11-09
3 Muse Ember Max Meta 25.0% 2026-11-09
4 Muse Wonder Max Meta 23.8% 2026-11-09
5 Muse Wonder Full Max Meta 23.3% 2026-11-09
6 Inkling 2 Small Thinking Machines 21.7% 2026-11-09
7 Grok 4.7 Extra High xAI 18.3% 2026-11-09
8 Gemini 4 Carbon High Google 17.5% 2026-11-09
9 Mistral Large 4 High Mistral 17.5% 2026-11-09
10 GLM-5.3 Max Zhipu AI 14.6% 2026-11-09
11 DeepSeek V4.1 Flash 1017 Max DeepSeek 14.2% 2026-11-09
12 MiMo-V2.6-Pro Xiaomi 13.8% 2026-11-09
13 MiMo-V2.6-Flash Xiaomi 13.3% 2026-11-09
14 Step 5 Preview StepFun 12.9% 2026-11-09
15 Kimi K3 Max Moonshot AI 12.9% 2026-11-09
16 Hy4 Tencent 12.5% 2026-11-09
17 Gemini 3.8 Flash High Google 12.1% 2026-11-09
18 Kimi K3.1 Pro Max Moonshot AI 11.7% 2026-11-09
19 DeepSeek V4 Pro 0813 Max DeepSeek 10.8% 2026-11-09
20 Claude Sonnet 5.5 Max Anthropic 10.4% 2026-11-09
21 GLM-5.3-Flash Max Zhipu AI 10.4% 2026-11-09
22 Claude Opus 5.5 Max Anthropic 9.2% 2026-11-09
23 Hy4 preview Tencent 8.8% 2026-11-09
24 MiniMax-M3.1-Flash Max MiniMax 7.5% 2026-11-09
25 GPT-6 Astra Max OpenAI 7.1% 2026-11-09
26 GPT-6.1 Sol Max OpenAI 6.3% 2026-11-09
27 Claude Fable 5.2 High Anthropic 5.0% 2026-11-09
28 Mercury 3 High Inception 5.0% 2026-11-09
29 Beam 1.0 Reflection 4.6% 2026-11-09
30 Claude Haiku 5.5 Max Anthropic 4.2% 2026-11-09
31 GPT-6 Luna Max OpenAI 4.2% 2026-11-09
32 Inkling Thinking Machines 3.8% 2026-11-09
33 Nemotron 3.5 Super VL 120B A12B High NVIDIA 3.8% 2026-11-09
34 Inkling Small Thinking Machines 2.9% 2026-11-09
35 GPT-6.1 Terra Max OpenAI 1.7% 2026-11-09