Instructions following Model Ranking

See which AI models perform best on Instructions following, which ones stay reliable, and where the biggest gaps appear.

Models Shown

Average Instructions following Score

8.6

Best Model

Failure Reasons

With failure reason Wrong answer61 With failure reason Did not follow instructions19 With failure reason Extra formatting3 With failure reason No answer2 With failure reason API error1 With failure reason Timed out1

216/216

Rank	Model	Company	Instructions following Score	Score	Total Cost	Tests Correct	Response Time (avg)
#214	Nemotron 3 Nano Omni 30b A3b Reasoning none	NVIDIA	4.8	3.2	$0.000	0/2	541ms
Total Tests 2 Wrong Tests 2 Total Cost $0.000 Response Time (avg) 541ms
#166	Laguna XS 2.1 none	Poolside	3.8	5.3	$0.008	0/2	364ms
Total Tests 2 Wrong Tests 2 Total Cost $0.008 Response Time (avg) 364ms
#178	MiniMax M2.7 medium	Minimax	3.8	5.0	$0.163	0/2	12.8s
Total Tests 2 Wrong Tests 2 Total Cost $0.163 Response Time (avg) 12.8s
#207	Granite 4.1 8B none	IBM Granite	3.6	4.0	$0.007	0/2	344ms
Total Tests 2 Wrong Tests 2 Total Cost $0.007 Response Time (avg) 344ms
#189	Trinity Large Preview none	Arcee AI	3.5	4.8	$0.008	0/2	822ms
Total Tests 2 Wrong Tests 2 Total Cost $0.008 Response Time (avg) 822ms
#209	Grok 4.1 Fast none	X AI	3.0	3.8	$0.008	0/2	685ms
Total Tests 2 Wrong Tests 2 Total Cost $0.008 Response Time (avg) 685ms

Instructions following Ranking