Released: 2026-05-30
Key:
Score Threshold:
google/gemini-2.0-flash-001
(75 * (100 / (5 + 1))) - 25
= 1225
moonshotai/kimi-k2.5
moonshot
((100 - 25) * 5 - 1) + 75 - 50
= 399
google/gemini-3-flash-preview
(25 * (50 - (100 / 75)))
= 1216
openai/gpt-4o
openai
((100 + 75) * 5) + (50 / 25)
= 877
deepseek/deepseek-chat-v3-0324
deepseek
((100 + 50) * (5 + 1)) + 25
= 925
anthropic/claude-sonnet-4.6
anthropic
((100 * 25) / 5) * (1 + 50 - 75)
= -12000
minimax/minimax-m2.5
minimax
100 * (75/25 + 5 + 1)
= 900
z-ai/glm-5
z-ai
(100 + 75) * 5 + 50 - 1
= 924
mistralai/mistral-medium-3
mistral
((100 - (50 / 25)) * (75 / 5)) + 1
= 1471
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.