Atlas

Benchmarks

← All benchmarks

HLAB-Vals-1.1 - Real Estate

Professional Work · 2026-10-19

Mean of two judges' task pass rates on the Real Estate practice-area tasks of HLAB-Vals-1.1.

Top models (higher is better)

ModelScore
Muse Wonder Full40.0
GPT-6.1 Sol30.0
Grok 4.730.0
Kimi K330.0
Muse Wonder30.0
Claude Haiku 5.520.0
Claude Sonnet 5.520.0
GLM-5.3 Flash20.0
MiMo-V2.6-Flash20.0
MiMo-V2.6-Pro20.0
Beam 1.0 501B A23B10.0
Claude Opus 5.510.0
DeepSeek V4.1 Flash 101710.0
DeepSeek V4 Pro 081310.0
Gemini 3.8 Flash10.0
Gemini 4 Argon10.0
Muse Spark Open10.0
GLM-5.30.0
GPT-6 Astra0.0
GPT-6 Luna0.0
Hy4 Preview0.0
Inkling0.0
Inkling-Small0.0
Mercury 30.0
MiniMax M3.1 Flash0.0
Mistral Large 4 Preview0.0
Loading Atlas data…