Puzzle Solving Model Ranking

See which AI models perform best on Puzzle Solving, which ones stay reliable, and where the biggest gaps appear. Sort by: Tests Correct ↓.

Models Shown

Average Puzzle Solving Score

0.7

Best Model

Gemini 3.6 Flash 1.0

Failure Reasons

With failure reason Wrong answer352 With failure reason Did not follow instructions166 With failure reason API error35 With failure reason Extra formatting16 With failure reason No answer14 With failure reason Timed out11

451/451

Rank	Model	Company	Puzzle Solving Score	Score	Total Cost	Tests Correct	Response Time (avg)
#232	Beam 1.0 501B A23B xhigh	Reflection	0.6	0.6	$0.688	1/3	4.46s
Total Tests 3 Wrong Tests 2 Total Cost $0.688 Response Time (avg) 4.46s
#236	Grok 4.3 medium	X AI	0.6	0.6	$0.989	1/3	22.5s
Total Tests 3 Wrong Tests 2 Total Cost $0.989 Response Time (avg) 22.5s
#237	Laguna XS 2.1 medium	Poolside	0.5	0.6	$0.083	1/3	3.43s
Total Tests 3 Wrong Tests 2 Total Cost $0.083 Response Time (avg) 3.43s
#244	Kimi K2.5 medium	Moonshot AI	0.5	0.6	$0.841	1/3	43.2s
Total Tests 3 Wrong Tests 2 Total Cost $0.841 Response Time (avg) 43.2s
#245	Beam 1.0 501B A23B low	Reflection	0.6	0.6	$0.432	1/3	3.42s
Total Tests 3 Wrong Tests 2 Total Cost $0.432 Response Time (avg) 3.42s
#248	MiMo-V2.5-Pro medium	Xiaomi	0.7	0.6	$0.332	1/3	5.31s
Total Tests 3 Wrong Tests 2 Total Cost $0.332 Response Time (avg) 5.31s
#249	Dots 3 Note Preview high	Dots Studio	0.6	0.6	$0.000	1/3	12.7s
Total Tests 3 Wrong Tests 2 Total Cost $0.000 Response Time (avg) 12.7s
#250	Dots 3 Note Preview low	Dots Studio	0.5	0.6	$0.000	1/3	8.52s
Total Tests 3 Wrong Tests 2 Total Cost $0.000 Response Time (avg) 8.52s
#253	Qwen3.6 27B none	Qwen	0.5	0.6	$0.148	1/3	5.15s
Total Tests 3 Wrong Tests 2 Total Cost $0.148 Response Time (avg) 5.15s
#263	DeepSeek V3.2 medium	DeepSeek	0.7	0.6	$0.152	1/3	37.7s
Total Tests 3 Wrong Tests 2 Total Cost $0.152 Response Time (avg) 37.7s
#264	Trinity Large Thinking low	Arcee AI	0.6	0.6	$0.918	1/3	3.28s
Total Tests 3 Wrong Tests 2 Total Cost $0.918 Response Time (avg) 3.28s
#268	KAT-Coder-Pro V2.5 medium	Kwaipilot	0.6	0.6	$0.669	1/3	3.20s
Total Tests 3 Wrong Tests 2 Total Cost $0.669 Response Time (avg) 3.20s
#277	Inkling none	Thinkingmachines	0.6	0.6	$0.292	1/3	931ms
Total Tests 3 Wrong Tests 2 Total Cost $0.292 Response Time (avg) 931ms
#278	Qwen3.5 Plus 2026-04-20 none	Qwen	0.7	0.6	$0.216	1/3	1.97s
Total Tests 3 Wrong Tests 2 Total Cost $0.216 Response Time (avg) 1.97s
#286	Inkling Small medium	Thinkingmachines	0.6	0.6	$0.113	1/3	5.24s
Total Tests 3 Wrong Tests 2 Total Cost $0.113 Response Time (avg) 5.24s

Puzzle Solving Ranking

Filter models

Top Models by Puzzle Solving Score

Puzzle Solving Score vs Total Cost

Top Models by Response Time (avg)