Released: 2026-03-21
Key:
Score Threshold:
google/gemini-2.0-flash-001
(100 - 10) * 7 + 50 - 5
= 675
deepseek/deepseek-chat-v3-0324
deepseek
(75 - 5) * (10 + 7)
= 1190
google/gemini-3-flash-preview
(100 + 10) * (7 - (50 / 75))
= 696
minimax/minimax-m2.5
minimax
(100 * 7) + (75 - 10) + 5
= 770
x-ai/grok-3-mini-beta
x-ai
(100 * 7) - (50 - 10)
= 660
z-ai/glm-5
z-ai
(100 * 7) - 50 + 10
= 660
moonshotai/kimi-k2.5
moonshot
7 * 100 - 50 + 10
= 660
openai/gpt-4o
openai
100 * 5 + 75 + (50 / 10)
= 580
anthropic/claude-3.7-sonnet
anthropic
(100 - 10) * 75 / 5
= 1350
anthropic/claude-sonnet-4.6
anthropic
75 * (10 - 5) + 100 + 50 + 7
= 532
mistralai/mistral-medium-3
mistral
(100 + 75 - 10) * (7 + 5)
= 1980
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.