Atlas

Benchmarks

← All benchmarks

Finance Agent v2.1 - Financial Modeling

Professional Work · 2026-10-19

Partial-credit score on the Financial Modeling category of Finance Agent v2.1.

Top models (higher is better)

ModelScore
GPT-6.1 Sol52.6
Claude Haiku 5.551.5
DeepSeek V4.1 Flash46.8
Gemini 3.8 Flash45.1
GPT-6 Astra45.0
GLM-5.342.3
MiMo-V2.6-Flash41.2
Muse Wonder Full40.6
DeepSeek V4.1 Flash 101739.6
MiMo-V2.6-Pro39.0
Muse Wonder38.7
Kimi K338.3
Grok 4.737.2
Gemini 4 Argon36.5
GPT-6 Luna36.5
Hy4 Preview30.3
Step 5 Preview29.3
DeepSeek V4 Pro 081327.2
MiniMax M3.1 Flash26.9
Inkling-Small25.0
Mistral Large 4 Preview24.5
Muse Spark Open23.4
GLM-5.3 Flash23.3
Beam 1.0 501B A23B20.8
Inkling19.7
Mercury 317.8
MiniMax M316.2
Loading Atlas data…