Released: 2026-05-12
Key:
Score Threshold:
openai/gpt-4o
openai
((75 + 5) * 7) + 25 - 100
= 485
google/gemini-3-flash-preview
(100 * 5) + 75 - (50 / (25 / 7))
= 561
anthropic/claude-sonnet-4.6
anthropic
75 * 7 + 50 - 25 + 100 / 5
= 570
deepseek/deepseek-chat-v3-0324
deepseek
((100 + 7) * 5) + 50 - 25
= 560
z-ai/glm-5
z-ai
7 * (75 + (100 + 50) / 25) + 5
= 572
x-ai/grok-3-mini-beta
x-ai
((100 * 5) + 75 - (7 - (50 / 25)))
= 570
google/gemini-2.0-flash-001
(100 * 5) + 75 + 7 - 50
= 532
minimax/minimax-m2.5
minimax
100 + 75 + 50 * 7 - 5
= 520
moonshotai/kimi-k2.5
moonshot
(100 - 50/25) * 5 + 75 + 7
= 572
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.