모델 순위

순위	변동	모델	점수	95% 구간	투표 수	가격 $/M
1	—	qwen3-235b-a22b-instruct-2507오픈Alibaba	1,422	±6	9,744	—
2	—	kimi-k2-0711-preview오픈Moonshot AI	1,421	±6	15,926	—
3	—	deepseek-r1-0528오픈DeepSeek · 변형 1개 더	1,417	±6	21,205	—
4	—	glm-4.5오픈Z.ai	1,413	±7	8,707	—
5	—	qwen3-235b-a22b-thinking-2507오픈Alibaba	1,403	±7	6,996	—
6	—	qwen3-235b-a22b-no-thinking오픈Alibaba	1,397	±5	27,765	—
7	—	deepseek-v3-0324오픈DeepSeek · 변형 1개 더	1,393	±5	30,936	—
8	—	qwen3-coder-480b-a35b-instruct오픈Alibaba	1,382	±6	9,603	—
9	—	qwen3-30b-a3b-instruct-2507오픈Alibaba	1,380	±7	6,951	—
10	—	glm-4.5-air오픈Z.ai	1,378	±7	8,075	—
11	—	qwen3-235b-a22b오픈Alibaba	1,375	±5	23,704	—
12	—	minimax-m1오픈MiniMax	1,370	±5	20,987	—
13	—	gemma-3-27b-it오픈Google	1,364	±4	35,379	—
14	—	gpt-oss-120b오픈OpenAI	1,354	±8	6,099	$0.26
15	▲ 4	step-3오픈StepFun	1,354	±11	2,973	—
16	▼ 1	mistral-small-2506오픈Mistral	1,353	±6	12,660	—
17	▼ 1	command-a-03-2025오픈Cohere	1,347	±4	34,458	—
18	▼ 1	qwen3-32b오픈Alibaba	1,346	±9	4,074	—
19	▼ 1	llama-3.1-nemotron-ultra-253b-v1오픈NVIDIA	1,345	±11	2,656	—
20	—	nvidia-llama-3.3-nemotron-super-49b-v1.5오픈NVIDIA	1,342	±9	3,917	—
21	—	gemma-3-12b-it오픈Google	1,341	±9	3,976	—
22	▲ 10	glm-4.5v오픈Z.ai	1,339	±18	992	—
23	▼ 1	qwq-32b오픈Alibaba · 변형 1개 더	1,337	±5	24,118	—
24	▼ 1	llama-3.1-405b-instruct-bf16오픈Meta	1,335	±4	43,788	—
25	▼ 1	llama-3.1-405b-instruct-fp8오픈Meta	1,333	±3	63,038	—
26	▼ 1	gpt-oss-20b오픈OpenAI	1,332	±8	5,721	—
27	▼ 1	llama-4-maverick-17b-128e-instruct오픈Meta	1,328	±5	28,106	—
28	▼ 1	qwen3-30b-a3b오픈Alibaba	1,327	±5	23,737	—
29	▼ 1	llama-3.3-nemotron-49b-super-v1오픈NVIDIA	1,325	±12	2,371	—
30	▼ 1	deepseek-v2.5-1210오픈DeepSeek · 변형 1개 더	1,322	±8	7,243	—
31	—	llama-3.3-70b-instruct오픈Meta	1,320	±3	54,132	—
32	▼ 2	llama-4-scout-17b-16e-instruct오픈Meta	1,319	±6	17,813	—
33	—	qwen-max-0919오픈Alibaba	1,317	±5	17,432	—
34	—	gemma-3n-e4b-it오픈Google	1,316	±6	16,502	—
35	—	athene-v2-chat오픈NexusFlow	1,314	±4	26,074	—
36	—	mistral-large-2407오픈Mistral · 변형 1개 더	1,313	±4	48,217	—
37	—	athene-70b-0725오픈NexusFlow	1,305	±5	20,580	—
38	—	gemma-3-4b-it오픈Google	1,303	±9	4,321	—
39	—	qwen2.5-72b-instruct오픈Alibaba	1,302	±4	41,519	—
40	—	llama-3.1-nemotron-70b-instruct오픈NVIDIA	1,298	±7	7,577	—
41	—	mistral-small-3.1-24b-instruct-2503오픈Mistral	1,296	±6	17,533	—
42	—	llama-3.1-70b-instruct오픈Meta	1,294	±3	58,637	—
43	—	llama-3.1-tulu-3-70b오픈Ai2	1,289	±10	3,010	—
44	—	jamba-1.5-large오픈AI21 Labs	1,289	±7	9,125	—
45	—	llama-3.1-nemotron-51b-instruct오픈NVIDIA	1,286	±10	3,889	—
46	—	gemma-2-27b-it오픈Google	1,286	±3	79,538	—
47	—	gemma-2-9b-it-simpo오픈Princeton	1,280	±7	10,548	—
48	—	nemotron-4-340b-instruct오픈NVIDIA	1,279	±5	20,608	—
49	—	command-r-plus-08-2024오픈Cohere	1,279	±6	10,535	—
50	—	llama-3-70b-instruct오픈Meta	1,276	±3	163,629	—
51	—	mistral-small-24b-instruct-2501오픈Mistral	1,275	±6	15,321	—
52	—	qwen2.5-coder-32b-instruct오픈Alibaba	1,271	±8	5,730	—
53	—	c4ai-aya-expanse-32b오픈Cohere	1,268	±5	28,768	—
54	—	command-r-plus오픈Cohere	1,265	±4	80,846	—
55	—	gemma-2-9b-it오픈Google	1,265	±4	57,197	—
56	—	qwen2-72b-instruct오픈Alibaba	1,264	±5	38,872	—
57	—	deepseek-coder-v2오픈DeepSeek	1,264	±6	15,753	—
58	—	phi-4오픈Microsoft	1,258	±4	25,213	—
59	—	olmo-2-0325-32b-instruct오픈Ai2	1,256	±11	3,460	—
60	—	command-r-08-2024오픈Cohere	1,255	±6	10,851	—
61	—	jamba-1.5-mini오픈AI21 Labs	1,241	±7	9,274	—
62	—	ministral-8b-2410오픈Mistral	1,240	±9	5,111	—
63	—	qwen1.5-110b-chat오픈Alibaba	1,237	±5	27,430	—
64	—	qwen1.5-72b-chat오픈Alibaba	1,236	±5	40,658	—
65	—	mixtral-8x22b-instruct-v0.1오픈Mistral	1,232	±4	53,751	—
66	—	command-r오픈Cohere	1,231	±5	56,398	—
67	—	c4ai-aya-expanse-8b오픈Cohere	1,228	±7	10,391	—
68	—	llama-3.1-tulu-3-8b오픈Ai2	1,225	±10	3,074	—
69	—	llama-3-8b-instruct오픈Meta	1,225	±4	109,056	—
70	—	yi-1.5-34b-chat오픈01 AI	1,217	±5	25,135	—
71	—	zephyr-orpo-141b-A35b-v0.1오픈HuggingFace	1,216	±10	4,854	—
72	—	llama-3.1-8b-instruct오픈Meta	1,215	±4	52,578	—
73	—	granite-3.1-8b-instruct오픈IBM	1,213	±10	3,289	—
74	—	qwen1.5-32b-chat오픈Alibaba	1,208	±6	22,765	—
75	—	phi-3-medium-4k-instruct오픈Microsoft	1,202	±5	26,105	—
76	▲ 1	gemma-2-2b-it오픈Google	1,201	±4	48,892	—
77	▼ 1	mixtral-8x7b-instruct-v0.1오픈Mistral	1,201	±4	76,126	—
78	—	internlm2_5-20b-chat오픈InternLM	1,199	±7	10,599	—
79	—	dbrx-instruct-preview오픈Databricks	1,198	±6	33,743	—
80	—	qwen1.5-14b-chat오픈Alibaba	1,196	±7	18,687	—
81	—	wizardlm-70b오픈Microsoft	1,188	±9	8,383	—
82	—	granite-3.0-8b-instruct오픈IBM	1,188	±8	7,002	—
83	—	yi-34b-chat오픈01 AI	1,187	±7	15,917	—
84	—	deepseek-llm-67b-chat오픈DeepSeek	1,186	±11	4,988	—
85	—	openchat-3.5-0106오픈OpenChat · 변형 1개 더	1,185	±8	12,990	—
86	—	granite-3.1-2b-instruct오픈IBM	1,184	±11	3,380	—
87	—	snowflake-arctic-instruct오픈Snowflake	1,183	±6	34,173	—
88	—	tulu-2-dpo-70b오픈AllenAI/UW	1,182	±10	6,658	—
89	—	gemma-1.1-7b-it오픈Google	1,180	±6	25,070	—
90	—	openhermes-2.5-mistral-7b오픈NousResearch	1,178	±10	5,088	—
91	—	vicuna-33b오픈LMSYS	1,177	±6	22,936	—
92	—	starling-lm-7b-beta오픈Nexusflow	1,177	±7	16,676	—
93	—	phi-3-small-8k-instruct오픈Microsoft	1,176	±6	18,476	—
94	—	llama-2-70b-chat오픈Meta	1,175	±5	39,595	—
95	—	llama-3.2-3b-instruct오픈Meta	1,172	±7	8,390	—
96	—	starling-lm-7b-alpha오픈UC Berkeley	1,172	±8	10,415	—
97	—	nous-hermes-2-mixtral-8x7b-dpo오픈NousResearch	1,170	±12	3,836	—
98	—	granite-3.0-2b-instruct오픈IBM	1,162	±8	7,191	—
99	—	llama2-70b-steerlm-chat오픈NVIDIA	1,162	±13	3,636	—
100	—	solar-10.7b-instruct-v1.0오픈Upstage AI	1,158	±13	4,286	—
101	—	dolphin-2.2.1-mistral-7b오픈Cognitive Computations	1,156	±15	1,714	—
102	—	mistral-7b-instruct-v0.2오픈Mistral	1,155	±6	20,067	—
103	—	mpt-30b-chat오픈MosaicML	1,153	±12	2,644	—
104	—	wizardlm-13b오픈Microsoft	1,153	±9	7,176	—
105	—	falcon-180b-chat오픈TII	1,148	±17	1,327	—
106	—	phi-3-mini-4k-instruct-june-2024오픈Microsoft	1,148	±6	12,808	—
107	—	qwen1.5-7b-chat오픈Alibaba	1,147	±10	4,872	—
108	—	llama-2-13b-chat오픈Meta	1,147	±7	19,722	—
109	—	vicuna-13b오픈LMSYS	1,145	±7	19,775	—
110	—	qwen-14b-chat오픈Alibaba	1,141	±11	5,065	—
111	—	codellama-34b-instruct오픈Meta	1,140	±9	7,509	—
112	▲ 1	gemma-7b-it오픈Google	1,136	±9	9,176	—
113	▼ 1	phi-3-mini-128k-instruct오픈Microsoft	1,136	±7	21,622	—
114	—	zephyr-7b-beta오픈HuggingFace	1,136	±9	11,321	—
115	—	phi-3-mini-4k-instruct오픈Microsoft	1,134	±6	21,097	—
116	—	zephyr-7b-alpha오픈HuggingFace	1,132	±16	1,811	—
117	—	guanaco-33b오픈UW	1,131	±12	2,996	—
118	—	smollm2-1.7b-instruct오픈HuggingFace	1,129	±13	2,375	—
119	—	codellama-70b-instruct오픈Meta	1,124	±18	1,192	—
120	—	stripedhyena-nous-7b오픈Together AI	1,123	±11	5,276	—
121	—	llama-3.2-1b-instruct오픈Meta	1,122	±7	8,523	—
122	—	vicuna-7b오픈LMSYS	1,118	±9	7,017	—
123	—	gemma-1.1-2b-it오픈Google	1,116	±7	11,351	—
124	—	llama-2-7b-chat오픈Meta	1,113	±7	14,532	—
125	—	mistral-7b-instruct오픈Mistral	1,113	±9	9,142	—
126	—	qwen1.5-4b-chat오픈Alibaba	1,094	±9	7,816	—
127	—	gemma-2b-it오픈Google	1,093	±11	4,918	—
128	—	olmo-7b-instruct오픈Ai2	1,080	±11	6,503	—
129	—	koala-13b오픈UC Berkeley	1,073	±10	7,020	—
130	—	gpt4all-13b-snoozy오픈Nomic AI	1,065	±15	1,788	—
131	—	alpaca-13b오픈Stanford	1,065	±11	5,864	—
132	—	mpt-7b-chat오픈MosaicML	1,063	±12	3,997	—
133	—	chatglm3-6b오픈Tsinghua	1,059	±12	4,763	—
134	—	RWKV-4-Raven-14B오픈RWKV	1,044	±11	4,920	—
135	—	chatglm2-6b오픈Tsinghua	1,029	±13	2,713	—
136	—	oasst-pythia-12b오픈OpenAssistant	1,023	±11	6,368	—
137	—	chatglm-6b오픈Tsinghua	999	±13	4,983	—
138	—	fastchat-t5-3b오픈LMSYS	993	±12	4,288	—
139	—	dolly-v2-12b오픈Databricks	978	±14	3,480	—
140	—	llama-13b오픈Meta	970	±16	2,446	—
141	—	stablelm-tuned-alpha-7b오픈Stability AI	955	±13	3,336	—

모델 91개 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