Released: 2026-03-24
Key:
Score Threshold:
google/gemini-2.0-flash-001
(50 * 4) * (4 + 3)
= 1400
deepseek/deepseek-chat-v3-0324
deepseek
((50 * (25 - 4)) + (3 * 3)) * 4
= 4236
anthropic/claude-3.7-sonnet
anthropic
4 * 3 * 50 + 4
= 604
openai/gpt-4o
openai
((50 + 25) * 4) * 3 + 4
= 904
mistralai/mistral-medium-3
mistral
((50 + 25) * (4 + 4)) * 3
= 1800
z-ai/glm-5
z-ai
50 * (25 - 3 * 3) + 4
= 804
x-ai/grok-3-mini-beta
x-ai
4 * (50 * 4 + (3 / 3))
= 804
moonshotai/kimi-k2.5
moonshot
50 * (25 - 3 * 3) + 4
= 804
google/gemini-3-flash-preview
(50 + 25 - 3) * (4 + 3 + 4)
= 792
minimax/minimax-m2.5
minimax
(50 * 4 + 3 / 3) * 4
= 804
anthropic/claude-sonnet-4.6
anthropic
(25 + 3) * (50 / 4 - 4) + 3
= 241
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.