General Intelligence x Wrong answer Ranking

See which AI models are most likely to hit Wrong answer on General Intelligence, so you can spot weak points faster. Sort by: Total Cost ↑.

Models Shown

Total Failures

Most Affected Model

Failure Reasons

Did not follow instructions117 Wrong answer95 API error14 Timed out4

Categories

Domain specific690 Anti-AI Tricks401 Coding374 Puzzle Solving287 Trivia259 Combined99 General Intelligence95 Instructions following86 Data parsing and extraction67 Tool Calling5

95/95

Rank	Model	Company	Wrong answer Count	Category Score	Total Cost	Tests Correct	Response Time (avg)
#111	Grok Build 0.1 medium	X AI	1	4.4	$1.130	0/1	18.4s
Total Tests 1 Wrong Tests 1 Total Cost $1.130 Response Time (avg) 18.4s
#155	Step 3.7 Flash high	Stepfun	1	5.5	$1.229	0/1	4.17s
Total Tests 1 Wrong Tests 1 Total Cost $1.229 Response Time (avg) 4.17s
#57	Grok 4.6 medium	X AI	1	4.1	$1.713	0/1	17.3s
Total Tests 1 Wrong Tests 1 Total Cost $1.713 Response Time (avg) 17.3s
#55	Grok 4.5 medium	X AI	1	6.5	$2.042	0/1	12.8s
Total Tests 1 Wrong Tests 1 Total Cost $2.042 Response Time (avg) 12.8s
#69	Qwen3.8 2.4T A95B low	Qwen	1	6.1	$2.672	0/1	23.0s
Total Tests 1 Wrong Tests 1 Total Cost $2.672 Response Time (avg) 23.0s

←

1 2 3 4 5 6 7

Filter models