General Intelligence x Did not follow instructions Ranking

See which AI models are most likely to hit Did not follow instructions on General Intelligence, so you can spot weak points faster.

Models Shown

Total Failures

115

Most Affected Model

Grok 4.5 1

Failure Reasons

Did not follow instructions115 Wrong answer90 API error14 Timed out4

Categories

Puzzle Solving128 General Intelligence115 Instructions following48 Anti-AI Tricks44 Coding29 Tool Calling12 Data parsing and extraction7 Domain specific2 Combined1

115/115

Rank	Model	Company	Did not follow instructions Count	Category Score	Total Cost	Tests Correct	Response Time (avg)
#24	Grok 4.5 high	X AI	1	4.7	$2.252	0/1	9.82s
Total Tests 1 Wrong Tests 1 Total Cost $2.252 Response Time (avg) 9.82s
#25	GPT-5.3-Codex medium	OpenAI	1	4.6	$0.988	0/1	4.87s
Total Tests 1 Wrong Tests 1 Total Cost $0.988 Response Time (avg) 4.87s
#27	Seed 2.1 Turbo high	Bytedance Seed	1	6.5	$1.797	0/1	25.7s
Total Tests 1 Wrong Tests 1 Total Cost $1.797 Response Time (avg) 25.7s
#37	Muse Spark 1.2 medium	Meta	1	6.1	$1.339	0/1	10.7s
Total Tests 1 Wrong Tests 1 Total Cost $1.339 Response Time (avg) 10.7s
#41	Qwen3.8 Max low	Qwen	1	6.5	$0.702	0/1	5.29s
Total Tests 1 Wrong Tests 1 Total Cost $0.702 Response Time (avg) 5.29s
#42	GPT-5.4 medium	OpenAI	1	4.7	$1.560	0/1	4.92s
Total Tests 1 Wrong Tests 1 Total Cost $1.560 Response Time (avg) 4.92s
#44	Muse Spark 1.2 low	Meta	1	5.0	$0.655	0/1	8.46s
Total Tests 1 Wrong Tests 1 Total Cost $0.655 Response Time (avg) 8.46s
#47	GPT-5.2 medium	OpenAI	1	3.7	$1.182	0/1	4.32s
Total Tests 1 Wrong Tests 1 Total Cost $1.182 Response Time (avg) 4.32s
#48	Qwen3.8 27B high	Qwen	1	5.0	~$0.287	0/1	8.45s
Total Tests 1 Wrong Tests 1 Total Cost ~$0.287 Response Time (avg) 8.45s
#49	Seed 2.1 Turbo medium	Bytedance Seed	1	6.1	$1.951	0/1	40.6s
Total Tests 1 Wrong Tests 1 Total Cost $1.951 Response Time (avg) 40.6s
#53	Gemini 2.5 Flash medium	Google	1	4.8	$0.644	0/1	4.86s
Total Tests 1 Wrong Tests 1 Total Cost $0.644 Response Time (avg) 4.86s
#55	Claude Sonnet 5 medium	Anthropic	1	4.8	$0.922	0/1	4.32s
Total Tests 1 Wrong Tests 1 Total Cost $0.922 Response Time (avg) 4.32s
#57	GPT-5 Mini medium	OpenAI	1	4.5	$0.241	0/1	13.5s
Total Tests 1 Wrong Tests 1 Total Cost $0.241 Response Time (avg) 13.5s
#67	GPT-5.2 Chat none	OpenAI	1	4.4	$0.594	0/1	3.20s
Total Tests 1 Wrong Tests 1 Total Cost $0.594 Response Time (avg) 3.20s
#69	Qwen3.8 27B low	Qwen	1	5.0	~$0.071	0/1	5.37s
Total Tests 1 Wrong Tests 1 Total Cost ~$0.071 Response Time (avg) 5.37s

Filter models

Top Models by Did not follow instructions Count

Did not follow instructions Count vs Score

Top Models by Response Time (avg)

Top Models by Estimated Wasted Cost

General Intelligence: Did not follow instructions

Filter models

Top Models by Did not follow instructions Count

Did not follow instructions Count vs Score

Top Models by Response Time (avg)

Top Models by Estimated Wasted CostEstimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 2.8999 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.286715.

Top Models by Estimated Wasted Cost