Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Puzzle Solving on PUZZLE Avg across difficulty levels
Loading...
54.3
Accuracy
GPT-5
32.98
38.515
44.05
49.585
Jun 2, 2026
Accuracy
Mean Completion Token Count
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Mean Completion Token Count
GPT-5
2026.06
54.3
12,867.2
DeepSeek V3.2
2026.06
43.8
23,038.1
Kimi K2
2026.06
43.8
36,343.4
Qwen 3 235B
Parameters=235B
2026.06
33.8
19,000.7
Feedback
Search any
task
Search any
task