	Claude 3.5 Haiku	Claude Sonnet 4	Command A	Deepseek Chat V3	Deepseek R1	Gemini 2.5 Flash	Gemini 2.5 Pro Preview 05 06	Mistral Large 2411	Mistral Medium 3	GPT 4.1	GPT 4.1 Mini	GPT 4.1 Nano	GPT 4o	GPT 4o Mini	O4 Mini	Grok 3	Grok 3 Mini	Grok 4
Score	12th 63.5%	3rd 75.9%	11th 63.7%	7th 72.5%	5th 73.5%	6th 73.1%	10th 67.2%	16th 51.6%	4th 75.6%	9th 68.6%	15th 56.9%	17th 49.6%	14th 59.0%	18th 48.6%	13th 61.8%	2nd 76.4%	8th 72.4%	1st 89.1%
64.6%	43%	78%	68%	70%	73%	78%	35%	58%	73%	55%	55%	43%	68%	60%	80%	70%	68%	88%
88.0%	95%	95%	85%	85%	98%	90%	43%	88%	90%	98%	88%	88%	90%	85%	90%	88%	90%	98%
69.3%	44%	100%	100%	85%	75%	67%	98%	40%	100%	100%	31%	56%	36%	6%	42%	100%	67%	100%
82.8%	75%	100%	100%	100%	100%	100%	100%	67%	92%	96%	38%	38%	100%	42%	42%	100%	100%	100%
60.3%	69%	72%	44%	69%	66%	75%	38%	38%	69%	56%	47%	50%	34%	28%	78%	81%	78%	94%
16.6%	3%	31%	3%	10%	16%	22%	16%	3%	7%	16%	6%	10%	10%	10%		47%	38%	34%
38.5%	34%	50%	33%	17%	59%	50%	67%	8%	25%	67%	13%	8%	8%	8%	17%	79%	54%	96%
35.8%	56%	52%	6%	40%	54%	75%	25%	21%	77%	13%	17%	19%	17%	15%	2%	17%	38%	100%
78.9%	50%	100%	100%	94%	100%	75%	100%	25%	75%	100%	100%	63%	100%	50%	50%	88%	50%	100%
70.2%	79%	92%	65%	71%	81%	83%	65%	58%	61%	69%	67%	56%	71%	65%	73%	58%	71%	79%
63.6%	92%	88%	25%	92%	25%	29%	96%	29%	96%	92%	29%	38%	46%	33%	38%	96%	100%	100%
71.1%	60%	60%	67%	69%	75%	73%	100%	44%	75%	81%	79%	48%	60%	48%	92%	77%	71%	100%
94.2%	92%	92%	100%	100%	100%	100%	71%	92%	100%	92%	92%	92%	96%	88%	92%	100%	96%	100%
75.6%	67%	100%	79%	77%	83%	98%	100%	67%	96%	54%	46%	52%	56%	50%	73%	90%	73%	100%
79.2%	85%	78%	75%	90%	78%	53%	85%	58%	83%	75%	88%	60%	75%	85%	85%	88%	90%	95%
67.6%	66%	63%	78%	69%	69%	88%	63%	59%	69%	63%	72%	63%	53%	56%	72%	66%	59%	88%
90.6%	83%	95%	95%	100%	100%	100%	65%	88%	100%	98%	95%	75%	73%	65%	98%	100%	100%	100%
75.3%	64%	75%	86%	81%	74%	79%	86%	65%	82%	59%	63%	84%	75%	57%	86%	72%	84%	84%
58.9%	45%	53%	60%	58%	73%	75%	58%	35%	53%	73%	55%	48%	43%	45%	65%	73%	53%	95%
64.6%	71%	71%	0%	71%	88%	67%	46%	67%	100%	55%	63%	0%	83%	79%	75%	71%	96%	59%
62.9%	56%	78%	69%	81%	63%	88%	72%	66%	75%	50%	63%	75%	56%	50%	0%	72%	66%	53%
54.3%	69%	47%	63%	66%	66%	44%	50%	59%	66%	47%	44%	25%	47%	44%	47%	47%	50%	97%

Model Similarity Dendrogram

Hierarchical clustering of models based on response similarity. Models grouped closer are more similar.