Released: 2026-06-26
Key:
Score Threshold:
anthropic/claude-sonnet-4.6
anthropic
((100 - 2) * 9) + 75 + 25 + 50
= 1032
deepseek/deepseek-chat-v3-0324
deepseek
((100 + 2) * 9) + 50 + 75 + 25
= 1068
openai/gpt-4o
openai
((75 + 25) * (9 + 2)) + 50 - 100
= 1050
minimax/minimax-m2.5
minimax
(100 + 50 + 2) * 9 + 25
= 1393
mistralai/mistral-medium-3
mistral
((100 - 9) * (75 + 2)) + 50 + 25
= 7082
z-ai/glm-5
z-ai
(100 + 2) * 9 + 50 - 75 / 25
= 965
google/gemini-3-flash-preview
(100 * 9) + 75 - (25 / (50 / 2))
= 974
Each model receives the same prompt with the numbers to use. Models are tasked with creating an expression using only arithmetic operations to reach the target number. Each number can only be used once and you do not have to use all the numbers. Their answers are evaluated without feedback or retries.