API error Failure Ranking

See which AI models run into API error most often, so you can spot reliability risks before choosing one. Sort by: Score ↑.

Models Shown

Total Failures

161

Most Affected Model

Categories

In category Coding45 In category Combined26 In category Tool Calling17 In category Anti-AI Tricks14 In category Data parsing and extraction14 In category Trivia13 In category General Intelligence12 In category Puzzle Solving12 In category Domain specific7 In category Instructions following1

68/68

Rank	Model	Company	API error Count	Score	Total Cost	Tests Correct	Response Time (avg)
#61	Qwen3.5 Plus 2026-02-15 medium	Qwen	1	7.5	$0.437	14/22	89.2s
Total Tests 22 Wrong Tests 8 Total Cost $0.437 Response Time (avg) 89.2s
#56	Kimi K2.7 Code medium	Moonshot AI	1	7.5	$0.740	12/22	84.2s
Total Tests 22 Wrong Tests 10 Total Cost $0.740 Response Time (avg) 84.2s
#55	Nemotron 3 Ultra medium	NVIDIA	1	7.5	$0.774	13/22	32.2s
Total Tests 22 Wrong Tests 9 Total Cost $0.774 Response Time (avg) 32.2s
#50	DeepSeek V4 Pro high	DeepSeek	1	7.7	$0.200	10/22	79.1s
Total Tests 22 Wrong Tests 12 Total Cost $0.200 Response Time (avg) 79.1s
#41	Qwen3.6 Plus medium	Qwen	1	7.8	$0.405	15/22	43.1s
Total Tests 22 Wrong Tests 7 Total Cost $0.405 Response Time (avg) 43.1s
#37	Kimi K3 max	Moonshot AI	2	8.0	$3.112	16/22	122.5s
Total Tests 22 Wrong Tests 6 Total Cost $3.112 Response Time (avg) 122.5s
#36	Inkling medium	Thinkingmachines	1	8.0	$0.391	15/22	16.2s
Total Tests 22 Wrong Tests 7 Total Cost $0.391 Response Time (avg) 16.2s
#30	Muse Spark 1.1 high	Meta	1	8.1	$1.694	12/22	31.5s
Total Tests 22 Wrong Tests 10 Total Cost $1.694 Response Time (avg) 31.5s

←

1 2 3 4 5

API error Failures