Released: 2026-03-19
Key:
Score Threshold:
mistralai/mistral-medium-3
mistral
((25 + 5) * (10 + 5)) - (10 * 5)
= 400
anthropic/claude-3.7-sonnet
anthropic
(10 * 5 * 5 + 25) * 10 / 5
= 550
deepseek/deepseek-chat-v3-0324
deepseek
(25 * (10 + 5)) + ((10 + 5) * 5)
= 450
google/gemini-2.0-flash-001
(10 * (25 + 10 + 5)) - 5 - 5
= 390
x-ai/grok-3-mini-beta
x-ai
(25 * (10 + 5 + (5 / 5)) + 10)
= 410
z-ai/glm-5
z-ai
(25 + 5 + 5 + 5) * 10 + 10
= 410
openai/gpt-4o
openai
(25 + 10 + 10) * 5 + 5 * 5
= 250
minimax/minimax-m2.5
minimax
(25+5+5+5)*10+10
= 410
anthropic/claude-sonnet-4.6
anthropic
25 * (10 + 5) + 10 + 5 - 5
= 385
moonshotai/kimi-k2.5
moonshot
((25 * 10) + (5 * 10)) + (5 * 5)
= 325
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.