Domain specific x Wrong answer Ranking

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

Models Shown

Total Failures

893

Most Affected Model

Failure Reasons

Wrong answer893 No answer66 Timed out46 API error40 Extra formatting36 Did not follow instructions3

Categories

Domain specific893 Anti-AI Tricks519 Coding484 Puzzle Solving352 Trivia315 General Intelligence134 Agentic128 Combined128 Instructions following117 Data parsing and extraction93

423/423

Rank	Model	Company	Wrong answer Count	Category Score	Total Cost	Tests Correct	Response Time (avg)
#58	Claude Fable 5 medium	Anthropic	1	0.5	$5.089	1/3	37.3s
Total Tests 3 Wrong Tests 2 Total Cost $5.089 Response Time (avg) 37.3s
#400	Grok 4.20 Multi Agent Beta medium	X AI	2	0.3	$5.599	0/3	24.7s
Total Tests 3 Wrong Tests 3 Total Cost $5.599 Response Time (avg) 24.7s
#8	GPT-6 Astra xhigh	OpenAI	1	0.8	$6.213	2/3	147.0s
Total Tests 3 Wrong Tests 1 Total Cost $6.213 Response Time (avg) 147.0s

Filter models