Released: 2026-04-06
Key:
Score Threshold:
google/gemini-2.0-flash-001
((75 * 10) - (100 + 50)) * (3 - 25)
= -13200
mistralai/mistral-medium-3
mistral
(75 + 50) * (100 - (10 + 3)) + 25
= 10900
openai/gpt-4o
openai
(75 * 10) - (25 + 3 - 50)
= 772
deepseek/deepseek-chat-v3-0324
deepseek
(100 * (75 - 3)) + (50 - 25 - 10)
= 7215
x-ai/grok-3-mini-beta
x-ai
((75 * 10) - (100 / (25 / 3)))
= 738
z-ai/glm-5
z-ai
(100 + 10) * (50 / 25 * 3) + 75
= 735
google/gemini-3-flash-preview
(75 * 10) - (50 / 100 * 25 + 3)
= 734
anthropic/claude-3.7-sonnet
anthropic
(75 - 3) * 10 * 100 / 25
= 2880
minimax/minimax-m2.5
minimax
(100+75+50+10)*3+25
= 730
moonshotai/kimi-k2.5
moonshot
(100 * 75 - 50 * 3) / 10
= 735
anthropic/claude-sonnet-4.6
anthropic
(75 * 10) - (50 / 25) + (100 - 3)
= 845
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.