Atlas

Benchmarks

← All benchmarks

Harvey HLAB - Criteria Pass Rate - Funds Asset Management

Professional Work · 2026-05-06

The Funds Asset Management practice-area split of Harvey's Legal Agent Benchmark held-out evaluation criteria pass rate. This reports pooled rubric criteria passed within that split and is distinct from the all-pass task-score split.

Top models (higher is better)

ModelScore
Gemini 4 Argon93.0
Muse Spark 1.192.1
Muse Wonder91.3
Muse Spark 1.291.0
Muse Spark 1.391.0
Muse Wonder Full90.7
MiMo-V2.6-Pro90.1
Gemini 3.8 Flash89.5
GLM-5.288.3
Step 5 Preview88.3
DeepSeek V4.1 Flash 101788.0
Grok 4.687.8
Grok 4.787.8
Mistral Large 4 Preview87.8
GLM-5.387.2
Claude Opus 586.6
Ember-186.6
Qwen3.8-Max86.6
DeepSeek-V4-Flash-073186.3
Grok 4.586.0
Hy4 Preview86.0
Kimi K385.7
Gemini 3.7 Flash85.1
Opus 4.784.5
Opus 4.884.5
GPT-6.1 Sol84.5
MiMo-V2.6-Flash84.0
Qwen3.8 27B84.0
GPT-6 Astra83.7
Sonnet 583.4
Sonnet 4.682.5
MiniMax M382.5
DeepSeek V4 Pro 081382.2
Kimi K2.682.2
Claude Haiku 5.581.9
Gemini 3.6 Flash81.3
DeepSeek V4.1 Flash81.0
MiMo-V2.5-Pro81.0
DeepSeek-V4-Pro80.8
Qwen3.7-Max (source unspecified)80.8
Loading Atlas data…