Released: 2026-07-21
Key:
Score Threshold:
openai/gpt-4o
openai
(6 * 75) + (50 + 4) - 100
= 404
google/gemini-3-flash-preview
(4 * (100 + 25 + 6)) + 50 + 75
= 649
mistralai/mistral-medium-3
mistral
((100 - 6) * (75 - (50 - 25))) + 4
= 4704
deepseek/deepseek-chat-v3-0324
deepseek
(100 * 6) - (75 - (50 + 25 + 4))
= 604
moonshotai/kimi-k2.5
moonshot
6*(100 - 75/25) - 50
= 532
minimax/minimax-m2.5
minimax
(100 + 75 - 50) * 4 + 6 + 25
= 531
anthropic/claude-sonnet-4.6
anthropic
((100 + 6) * 4) + 75 + 25 + 50
= 574
z-ai/glm-5
z-ai
((100 + 25) * 4) + 75 - 50 + 6
= 531
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.