AI vs Puzzles

Back to Home
Puzzle#128

Countdown

Released: 2026-03-04


11
Models Tested
5
Solved
9
LLM Providers

Key:

Used
Unused

Score Threshold:

Within 7

deepseek/deepseek-chat-v3-0324

deepseek

202Target
75
25
100
9
1
10

(100 + 75) * (9 - 1) - 25 + 10

= 1385

Failed

openai/gpt-4o

openai

202Target
75
25
100
9
1
10

(25 * 9) + 1 + (100 / 10)

= 236

Failed

anthropic/claude-3.7-sonnet

anthropic

202Target
75
25
100
9
1
10

75 * 10 / 25 * 100 - 9 - 1

= 2990

Failed

Minimax

minimax/minimax-m2.5

minimax

202Target
75
25
100
9
1
10

(100 + 75) + (25 + 10 - 9 + 1)

= 202

Perfect Solution!

x-ai/grok-3-mini-beta

x-ai

202Target
75
25
100
9
1
10

(100 + 75 + 25) + (10 - 9 + 1)

= 202

Perfect Solution!

google/gemini-2.0-flash-001

google

202Target
75
25
100
9
1
10

(100 - 75) * (10 - 1) - 25 + 9

= 209

Failed

mistralai/mistral-medium-3

mistral

202Target
75
25
100
9
1
10

((100 + 75) + (25 * (10 - (9 + 1))))

= 175

Failed

MoonshotAI

moonshotai/kimi-k2.5

moonshot

202Target
75
25
100
9
1
10

(75 + 25) + 100 + 10 - 9 + 1

= 202

Perfect Solution!

anthropic/claude-sonnet-4.6

anthropic

202Target
75
25
100
9
1
10

100 + 75 + 25 + 1 + 10 - 9

= 202

Perfect Solution!

google/gemini-3-flash-preview

google

202Target
75
25
100
9
1
10

(100 * (25 - 9 - 1)) / 75 + 10

= 30

Failed

Z.ai

z-ai/glm-5

z-ai

202Target
75
25
100
9
1
10

100 + 75 + 25 + 10 - 9 + 1

= 202

Perfect Solution!

Methodology Note

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.

Leaderboard