Released: 2026-04-01
Key:
Score Threshold:
google/gemini-2.0-flash-001
(75 * 10) + 50 + 2 * 2 - 4
= 800
openai/gpt-4o
openai
(75 * 10) + (4 * 50) - 2
= 948
mistralai/mistral-medium-3
mistral
((75 + 50) * (10 + 2)) - (4 / 2)
= 1498
anthropic/claude-3.7-sonnet
anthropic
(75 - 2) * (10 + 2)
= 876
deepseek/deepseek-chat-v3-0324
deepseek
((75 + 50) * (10 - 2)) + (4 + 2)
= 1006
minimax/minimax-m2.5
minimax
(75 * (10 + 2)) - (50 / (4 / 2))
= 875
x-ai/grok-3-mini-beta
x-ai
(75 * (10 + 2) - (4 * 2))
= 892
z-ai/glm-5
z-ai
50 * 4 * (2 + 2) + 75 + 10
= 885
google/gemini-3-flash-preview
(10 * (75 + 50 / 4)) - 2
= 873
moonshotai/kimi-k2.5
moonshot
(50*2-4)*10-75
= 885
anthropic/claude-sonnet-4.6
anthropic
50 * (2 + 10 + 4 + 2) - 75
= 825
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.