Atlas

Benchmarks

← All benchmarks

Finance Agent v2.1 - Market Analysis

Professional Work · 2026-10-19

Partial-credit score on the Market Analysis category of Finance Agent v2.1.

Top models (higher is better)

ModelScore
Gemini 4 Argon70.7
Gemini 3.8 Flash70.5
GLM-5.3 Flash67.9
Muse Wonder67.7
Muse Wonder Full67.4
DeepSeek V4.1 Flash67.1
Hy4 Preview66.9
Mistral Large 4 Preview66.6
Gemini 4 Carbon66.0
Step 5 Preview65.2
GPT-6 Astra64.2
Grok 4.764.0
DeepSeek V4 Pro 081363.0
MiniMax M362.2
Claude Haiku 5.561.4
Qwen3.8 Max (0902)61.1
MiMo-V2.6-Pro61.0
GLM-5.361.0
GPT-6.1 Sol61.0
DeepSeek V4.1 Flash 101760.7
Mercury 360.7
Kimi K3.1 Pro59.4
Muse Spark Open56.6
GPT-6.1 Terra56.3
MiniMax M3.1 Flash55.7
MiMo-V2.6-Flash55.0
GPT-6 Luna52.4
Kimi K352.1
Beam 1.0 501B A23B50.9
Inkling49.8
Inkling-Small43.6
Loading Atlas data…