모델 순위

순위	변동	모델	지능	가격 $/M	속도 tok/s	첫 토큰
1	—	Qwen3 235B A22B (Reasoning)오픈Alibaba	62.3	—	—	—
2	—	Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)오픈NVIDIA	60.8	—	—	—
3	—	DeepSeek R1오픈DeepSeek	60.2	—	—	—
4	—	Qwen3 32B (Reasoning)오픈Alibaba	59.2	—	—	—
5	—	QwQ 32B오픈Alibaba	58.1	—	—	—
6	—	Qwen3 14B (Reasoning)오픈Alibaba	55.7	—	—	—
7	—	Qwen3 30B A3B (Reasoning)오픈Alibaba	55.6	—	—	—
8	—	DeepSeek V3 0324 (Mar' 25)오픈DeepSeek	53.2	—	—	—
9	—	DeepSeek R1 Distill Qwen 32B오픈DeepSeek	51.9	—	—	—
10	—	Qwen3 8B (Reasoning)오픈Alibaba	51.4	—	—	—
11	—	Llama 3.3 Nemotron Super 49B v1 (Reasoning)오픈NVIDIA	51.2	—	—	—
12	—	Llama 4 Maverick오픈Meta	50.5	—	—	—
13	—	DeepSeek R1 Distill Qwen 14B오픈DeepSeek	49.0	—	—	—
14	—	DeepSeek R1 Distill Llama 70B오픈DeepSeek	48.2	—	—	—
15	—	Qwen3 4B (Reasoning)오픈Alibaba	47.3	—	—	—
16	—	Reka Flash 3오픈Reka AI	47.1	—	—	—
17	—	Qwen3 235B A22B오픈Alibaba	47.1	—	—	—
18	—	DeepSeek V3 (Dec '24)오픈DeepSeek	45.6	—	—	—
19	—	Qwen3 32B오픈Alibaba	43.5	—	—	—
20	—	Llama 4 Scout오픈Meta	43.0	—	—	—
21	—	QwQ 32B-Preview오픈Alibaba	42.6	—	—	—
22	—	Qwen3 30B A3B오픈Alibaba	42.6	—	—	—
23	—	Llama 3.3 Instruct 70B오픈Meta	41.1	—	—	—
24	—	Qwen3 14B오픈Alibaba	41.0	—	—	—
25	—	Llama 3.1 Instruct 405B오픈Meta	40.5	—	—	—
26	—	Qwen2.5 Instruct 72B오픈Alibaba	40.4	—	—	—
27	—	MiniMax-Text-01오픈MiniMax	40.2	—	—	—
28	—	Phi-4오픈Microsoft Azure	40.2	—	—	—
29	—	Command A오픈Cohere	40.0	—	—	—
30	—	Llama 3.1 Tulu3 405B오픈Allen Institute for AI	39.9	—	—	—
31	—	Llama 3.3 Nemotron Super 49B v1오픈NVIDIA	39.3	—	—	—
32	—	Mistral Large 2 (Nov '24)오픈Mistral	38.3	—	—	—
33	—	Qwen3 1.7B (Reasoning)오픈Alibaba	38.2	—	—	—
34	—	Gemma 3 27B Instruct오픈Google	37.6	—	—	—
35	—	Pixtral Large오픈Mistral	37.4	—	—	—
36	—	Qwen2.5 Instruct 32B오픈Alibaba	37.4	—	—	—
37	—	Llama 3.1 Nemotron Instruct 70B오픈NVIDIA	37.3	—	—	—
38	—	Qwen3 8B오픈Alibaba	36.7	—	—	—
39	—	Mistral Large 2 (Jul '24)오픈Mistral	36.6	—	—	—
40	—	Qwen2.5 Coder Instruct 32B오픈Alibaba	36.3	—	—	—
41	—	Llama 3.1 Instruct 70B오픈Meta	35.4	—	—	—
42	—	Mistral Small 3.1오픈Mistral	35.3	—	—	—
43	—	Mistral Small 3오픈Mistral	35.3	—	—	—
44	—	Qwen3 4B오픈Alibaba	35.2	—	—	—
45	—	DeepSeek R1 Distill Llama 8B오픈DeepSeek	34.0	—	—	—
46	—	Gemma 3 12B Instruct오픈Google	33.8	—	—	—
47	—	Llama 3.2 Instruct 90B (Vision)오픈Meta	33.4	—	—	—
48	—	Qwen2 Instruct 72B오픈Alibaba	32.6	—	—	—
49	—	DeepHermes 3 - Mistral 24B Preview오픈Nous Research	30.0	—	—	—
50	—	Jamba 1.5 Large오픈AI21 Labs	29.3	—	—	—
51	—	Hermes 3 - Llama-3.1 70B오픈Nous Research	29.1	—	—	—
52	—	Jamba 1.6 Large오픈AI21 Labs	28.8	—	—	—
53	—	Llama 3 Instruct 70B오픈Meta	27.5	—	—	—
54	—	Mistral Small (Sep '24)오픈Mistral	27.4	—	—	—
55	—	Phi-4 Multimodal Instruct오픈Microsoft Azure	26.9	—	—	—
56	—	Qwen2.5 Coder Instruct 7B오픈Alibaba	26.7	—	—	—
57	—	Mixtral 8x22B Instruct오픈Mistral	26.2	—	—	—
58	—	Phi-4 Mini Instruct오픈Microsoft Azure	26.0	—	—	—
59	—	Qwen3 1.7B오픈Alibaba	25.0	—	—	—
60	—	Phi-3 Medium Instruct 14B오픈Microsoft Azure	24.5	—	—	—
61	—	Gemma 3 4B Instruct오픈Google	24.5	—	—	—
62	—	Llama 3.1 Instruct 8B오픈Meta	23.7	—	—	—
63	—	Pixtral 12B (2409)오픈Mistral	23.4	—	—	—
64	—	Qwen3 0.6B (Reasoning)오픈Alibaba	23.2	—	—	—
65	—	Ministral 8B오픈Mistral	22.3	—	—	—
66	—	Gemma 2 9B오픈Google	22.2	—	—	—
67	—	Phi-3 Mini Instruct 3.8B오픈Microsoft Azure	22.1	—	—	—
68	—	Command-R+ (Aug '24)오픈Cohere	21.5	—	—	—
69	—	Llama 3 Instruct 8B오픈Meta	21.5	—	—	—
70	—	Codestral (May '24)오픈Mistral	20.4	—	—	—
71	—	Aya Expanse 32B오픈Cohere	20.1	—	—	—
72	—	Llama 2 Chat 13B오픈Meta	19.9	—	—	—
73	—	Command-R+ (Apr '24)오픈Cohere	19.9	—	—	—
74	—	DBRX Instruct오픈Databricks	19.8	—	—	—
75	—	Mistral NeMo오픈Mistral	19.6	—	—	—
76	—	Llama 3.2 Instruct 3B오픈Meta	19.5	—	—	—
77	—	DeepSeek R1 Distill Qwen 1.5B오픈DeepSeek	19.2	—	—	—
78	—	Jamba 1.5 Mini오픈AI21 Labs	18.4	—	—	—
79	—	Jamba 1.6 Mini오픈AI21 Labs	17.7	—	—	—
80	—	Mixtral 8x7B Instruct오픈Mistral	17.0	—	—	—
81	—	Qwen3 0.6B오픈Alibaba	16.7	—	—	—
82	—	DeepHermes 3 - Llama-3.1 8B Preview오픈Nous Research	16.2	—	—	—
83	—	Aya Expanse 8B오픈Cohere	16.0	—	—	—
84	—	Command-R (Aug '24)오픈Cohere	14.8	—	—	—
85	—	Command-R (Mar '24)오픈Cohere	14.7	—	—	—
86	—	Codestral-Mamba오픈Mistral	13.9	—	—	—
87	—	Mistral 7B Instruct오픈Mistral	10.1	—	—	—
88	—	Llama 3.2 Instruct 1B오픈Meta	9.7	—	—	—
89	—	Llama 2 Chat 7B오픈Meta	8.3	—	—	—

모델 39개 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