Released: 2026-04-19
Key:
Score Threshold:
google/gemini-2.0-flash-001
(100 * 3) + (8 - 3 - 3 - 7)
= 295
google/gemini-3-flash-preview
3 * (100 + (3 * 3) - 7)
= 306
mistralai/mistral-medium-3
mistral
(100 + (7 * 3)) * (3 + 3)
= 726
anthropic/claude-3.7-sonnet
anthropic
100 * 3 + 3 + 3
= 306
anthropic/claude-sonnet-4.6
anthropic
3 * 100 + 8 - 3 + 7 - 3
= 309
z-ai/glm-5
z-ai
100 * 3 + 3 + 3
= 306
deepseek/deepseek-chat-v3-0324
deepseek
(100 + 8) * 3
= 324
x-ai/grok-3-mini-beta
x-ai
(100 * 3) + 3 + 3
= 306
minimax/minimax-m2.5
minimax
(100*3)+(7-3/3)
= 306
moonshotai/kimi-k2.5
moonshot
100 * 3 + 3 + 3
= 306
openai/gpt-4o
openai
(3 * 100) + 8 + 3 + 3
= 314
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.