Released: 2026-06-12
Key:
Score Threshold:
z-ai/glm-5
z-ai
((100 + 4) * 4) + 5 + 10 - 3
= 428
mistralai/mistral-medium-3
mistral
(100 * (5 - (4 / 4))) + (10 * 3)
= 430
openai/gpt-4o
openai
(100 * 5) - (4 * 3) - 4
= 484
deepseek/deepseek-chat-v3-0324
deepseek
(100 - (5 * 3)) * (10 - 4) + 4
= 514
minimax/minimax-m2.5
minimax
100 * 5 - 10 - 3 - 4 - 4
= 479
moonshotai/kimi-k2.5
moonshot
(100 - 4) * 5 - 3
= 477
google/gemini-3-flash-preview
(5 * (100 - 4)) - (3 + 10) / 4
= 476
anthropic/claude-sonnet-4.6
anthropic
100 * 5 - (10 + 4 + 4 - 3)
= 485
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.