모델 순위

순위	변동	모델	지능	가격 $/M	속도 tok/s	첫 토큰
1	—	Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)클로즈드Anthropic	59.9	$20	60	103초
2	—	GPT-5.6 Sol (max)클로즈드OpenAI · 변형 5개 더	58.9	$11	72	170초
3	—	Claude Opus 4.8 (Adaptive Reasoning, Max Effort)클로즈드Anthropic	55.7	$10	56	46초
4	—	GPT-5.6 Terra (max)클로즈드OpenAI · 변형 5개 더	55.0	$5.6	141	155초
5	—	GPT-5.5 (xhigh)클로즈드OpenAI · 변형 4개 더	54.8	$11	69	98초
6	—	Claude Opus 4.7 (Adaptive Reasoning, Max Effort)클로즈드Anthropic · 변형 1개 더	53.5	$10	47	23초
7	—	Claude Sonnet 5 (Adaptive Reasoning, Max Effort)클로즈드Anthropic · 변형 1개 더	53.4	$4.0	77	221초
8	—	GPT-5.4 (xhigh)클로즈드OpenAI · 변형 2개 더	51.4	$5.6	156	113초
9	—	GPT-5.6 Luna (max)클로즈드OpenAI · 변형 5개 더	51.2	$2.3	217	99초
10	—	Muse Spark 1.1 (xhigh)클로즈드Meta	50.6	$2.0	119	1.1초
11	—	Gemini 3.5 Flash (high)클로즈드Google · 변형 2개 더	50.2	$3.4	152	24초
12	—	Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)클로즈드Anthropic · 변형 2개 더	47.2	$6.0	66	164초
13	—	Gemini 3.1 Pro Preview클로즈드Google	46.5	$4.5	128	25초
14	—	GPT-5.3 Codex (xhigh)클로즈드OpenAI	44.3	$4.8	91	75초
15	—	Claude Opus 4.6 (Adaptive Reasoning, Max Effort)클로즈드Anthropic · 변형 1개 더	43.7	$10	43	26초
16	—	Muse Spark클로즈드Meta	43.1	—	—	—
17	—	GPT-5.2 (xhigh)클로즈드OpenAI · 변형 2개 더	42.2	$4.8	74	150초
18	—	Claude Opus 4.5 (Reasoning)클로즈드Anthropic · 변형 1개 더	40.8	$10	48	17초
19	—	GPT-5.2 Codex (xhigh)클로즈드OpenAI	40.1	$4.8	160	74초
20	—	GPT-5.4 mini (xhigh)클로즈드OpenAI · 변형 2개 더	40.0	$1.7	164	9.2초
21	—	Gemini 3 Pro Preview (high)클로즈드Google · 변형 1개 더	39.6	$4.5	—	—
22	—	GPT-5.4 nano (xhigh)클로즈드OpenAI · 변형 2개 더	38.2	$0.46	171	7.4초
23	—	Gemini 3 Flash Preview (Reasoning)클로즈드Google · 변형 1개 더	37.8	$1.1	172	6.9초
24	—	GPT-5.1 (high)클로즈드OpenAI · 변형 1개 더	36.9	$3.4	101	40초
25	—	Claude 4.5 Sonnet (Reasoning)클로즈드Anthropic · 변형 1개 더	36.4	$6.0	41	14초
26	—	GPT-5 Codex (high)클로즈드OpenAI	36.1	$3.4	172	8.5초
27	—	GPT-5.1 Codex (high)클로즈드OpenAI	34.7	$3.4	167	7.7초
28	—	GPT-5 (high)클로즈드OpenAI · 변형 4개 더	34.7	$3.4	87	81초
29	—	Claude 4.1 Opus (Reasoning)클로즈드Anthropic · 변형 1개 더	33.7	$30	29	15초
30	—	GPT-5.5 Instant (May 2026)클로즈드OpenAI · 변형 1개 더	33.5	$11	—	—
31	—	o3-pro클로즈드OpenAI	32.5	$35	36	68초
32	—	Claude 4 Opus (Reasoning)클로즈드Anthropic · 변형 1개 더	31.0	$30	—	—
33	—	GPT-5 mini (medium)클로즈드OpenAI · 변형 2개 더	30.9	$0.69	105	17초
34	—	GPT-5.1 Codex mini (high)클로즈드OpenAI	30.6	$0.69	207	14초
35	—	o3클로즈드OpenAI	30.4	$3.5	123	6.0초
36	—	Claude 4.5 Haiku (Reasoning)클로즈드Anthropic · 변형 1개 더	29.6	$2.0	102	19초
37	—	Claude 4 Sonnet (Reasoning)클로즈드Anthropic · 변형 1개 더	28.9	$6.0	—	—
38	—	Claude 3.7 Sonnet (Reasoning)클로즈드Anthropic · 변형 1개 더	27.1	—	—	—
39	—	Gemini 2.5 Pro클로즈드Google	25.8	$3.4	139	19초
40	—	o4-mini (high)클로즈드OpenAI	25.6	$1.9	165	23초
41	—	Mercury 2클로즈드Inception	25.3	$0.38	783	3.0초
42	—	Gemini 3.1 Flash-Lite클로즈드Google	25.0	$0.56	272	5.8초
43	—	Gemini 2.5 Flash Preview (Sep '25) (Reasoning)클로즈드Google · 변형 3개 더	23.8	—	—	—
44	—	o1클로즈드OpenAI · 변형 1개 더	23.4	$26	101	30초
45	—	Gemini 2.5 Pro Preview (Mar' 25)클로즈드Google · 변형 1개 더	23.0	—	—	—
46	—	Nova 2.0 Pro Preview (medium)클로즈드Amazon · 변형 2개 더	21.8	$3.4	126	15초
47	—	Nova 2.0 Omni (medium)클로즈드Amazon · 변형 2개 더	20.9	$0.85	—	—
48	—	Gemini 2.5 Flash (Reasoning)클로즈드Google · 변형 1개 더	20.1	$0.85	211	15초
49	—	GPT-5 nano (high)클로즈드OpenAI · 변형 2개 더	19.9	$0.14	152	101초
50	—	GPT-4.1클로즈드OpenAI	19.4	$3.5	109	0.9초
51	—	Nova 2.0 Lite (medium)클로즈드Amazon · 변형 3개 더	19.0	$0.85	149	17초
52	—	o3-mini클로즈드OpenAI · 변형 1개 더	19.0	$1.9	226	6.2초
53	—	o1-pro클로즈드OpenAI	18.9	$263	—	—
54	—	Sonar Reasoning Pro클로즈드Perplexity	17.8	—	—	—
55	—	Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning)클로즈드Google · 변형 1개 더	15.1	$0.18	—	—
56	—	GPT-4.1 mini클로즈드OpenAI	14.8	$0.70	78	0.9초
57	—	o1-mini클로즈드OpenAI	14.0	—	—	—
58	—	GPT-4.5 (Preview)클로즈드OpenAI	13.6	—	—	—
59	—	Gemini 2.0 Flash Thinking Experimental (Jan '25)클로즈드Google · 변형 1개 더	13.3	—	—	—
60	—	Nova Premier클로즈드Amazon	12.7	$5.0	35	2.9초
61	—	GPT-4o (March 2025, chatgpt-4o-latest)클로즈드OpenAI · 변형 4개 더	12.3	—	—	—
62	—	Gemini 2.0 Flash (Feb '25)클로즈드Google · 변형 1개 더	12.3	$0.26	—	—
63	—	Claude 3.5 Haiku클로즈드Anthropic	12.3	$1.6	—	—
64	—	Gemini 2.0 Pro Experimental (Feb '25)클로즈드Google	11.8	—	—	—
65	—	Claude 3 Opus클로즈드Anthropic	11.8	$30	—	—
66	—	Sonar Reasoning클로즈드Perplexity	11.7	—	—	—
67	—	Gemini 2.5 Flash-Lite (Reasoning)클로즈드Google · 변형 1개 더	11.4	$0.18	223	26초
68	—	Gemini 1.5 Pro (Sep '24)클로즈드Google · 변형 1개 더	10.0	—	—	—
69	—	Claude 3.5 Sonnet (Oct '24)클로즈드Anthropic · 변형 1개 더	9.9	$6.0	—	—
70	—	GPT-4.1 nano클로즈드OpenAI	9.6	$0.18	154	0.7초
71	—	Sonar클로즈드Perplexity	9.5	—	—	—
72	—	Sonar Pro클로즈드Perplexity	9.3	—	—	—
73	—	Gemini 2.0 Flash-Lite (Feb '25)클로즈드Google · 변형 1개 더	8.8	—	—	—
74	—	Gemini 1.5 Flash (Sep '24)클로즈드Google · 변형 1개 더	8.0	—	—	—
75	—	GPT-4 Turbo클로즈드OpenAI	7.9	$15	32	2.8초
76	—	Nova Pro클로즈드Amazon	7.7	$1.4	—	—
77	—	GPT-4클로즈드OpenAI	7.0	$38	—	—
78	—	Nova Lite클로즈드Amazon	6.9	$0.11	181	1.0초
79	—	GPT-4o mini클로즈드OpenAI	6.9	$0.26	67	0.9초
80	—	Gemini 1.5 Flash-8B클로즈드Google	5.5	—	—	—
81	—	Claude 3 Sonnet클로즈드Anthropic	4.7	$6.0	—	—
82	—	Nova Micro클로즈드Amazon	4.7	$0.061	290	1.1초
83	—	Gemini 1.0 Ultra클로즈드Google	4.6	—	—	—
84	—	Claude 2.1클로즈드Anthropic	3.9	—	—	—
85	—	Claude 3 Haiku클로즈드Anthropic	3.9	$0.50	—	—
86	—	Claude 2.0클로즈드Anthropic	3.6	—	—	—
87	—	GPT-3.5 Turbo클로즈드OpenAI	3.6	$0.75	—	—
88	—	LFM 40B클로즈드Liquid AI	3.4	—	—	—
89	—	PALM-2클로즈드Google	3.2	—	—	—
90	—	Gemini 1.0 Pro클로즈드Google	3.1	—	—	—
91	—	Claude Instant클로즈드Anthropic	2.1	—	—	—

모델 41개 더 보기

지능의 발전

각 개발사가 그 시점에 가진 최고 모델의 Artificial Analysis Intelligence Index를 모델 공개 월 기준으로 보여 줍니다.

가장 많이 오름 · Anthropic +57.2현재 최고 · Anthropic 64.6최고와 최저의 차이 18.1점기간 2025년 11월 – 2026년 11월

64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi