Model rankings

Rank	Change	Model	Score	95% interval	Votes	Price $/M
1	—	deepseek-r1OpenDeepSeek	1,320	±7	8,580	—
2	—	deepseek-v3OpenDeepSeek	1,276	±4	22,007	—
3	—	llama-3.1-405b-instruct-bf16OpenMeta	1,254	±3	34,228	—
4	—	llama-3.1-405b-instruct-fp8OpenMeta	1,253	±3	63,136	—
5	—	deepseek-v2.5-1210OpenDeepSeek · 1 more variant	1,240	±6	7,247	—
6	—	llama-3.3-70b-instructOpenMeta	1,235	±4	28,572	—
7	—	mistral-large-2407OpenMistral · 1 more variant	1,232	±3	48,180	—
8	—	qwen-max-0919OpenAlibaba	1,231	±5	17,429	—
9	—	athene-v2-chatOpenNexusFlow	1,231	±5	26,092	—
10	—	athene-70b-0725OpenNexusFlow	1,223	±4	20,600	—
11	—	qwen2.5-72b-instructOpenAlibaba	1,222	±3	41,526	—
12	—	llama-3.1-nemotron-70b-instructOpenNVIDIA	1,216	±7	7,602	—
13	—	llama-3.1-70b-instructOpenMeta	1,214	±3	58,746	—
14	—	llama-3.1-tulu-3-70bOpenAi2	1,208	±11	3,011	—
15	—	gemma-2-27b-itOpenGoogle	1,207	±3	78,930	—
16	—	jamba-1.5-largeOpenAI21 Labs	1,205	±6	9,123	—
17	—	llama-3.1-nemotron-51b-instructOpenNVIDIA	1,204	±8	3,890	—
18	—	gemma-2-9b-it-simpoOpenPrinceton	1,200	±5	10,553	—
19	—	command-r-plus-08-2024OpenCohere	1,197	±6	10,539	—
20	—	nemotron-4-340b-instructOpenNVIDIA	1,196	±5	20,604	—
21	—	llama-3-70b-instructOpenMeta	1,195	±3	163,706	—
22	—	qwen2.5-coder-32b-instructOpenAlibaba	1,190	±7	5,730	—
23	▲ 1	mistral-small-24b-instruct-2501OpenMistral	1,186	±5	9,070	—
24	▼ 1	c4ai-aya-expanse-32bOpenCohere	1,185	±4	28,744	—
25	—	command-r-plusOpenCohere	1,183	±3	80,826	—
26	—	deepseek-coder-v2OpenDeepSeek	1,182	±5	15,755	—
27	—	qwen2-72b-instructOpenAlibaba	1,181	±3	38,867	—
28	—	gemma-2-9b-itOpenGoogle	1,180	±3	56,577	—
29	—	command-r-08-2024OpenCohere	1,176	±5	10,833	—
30	—	phi-4OpenMicrosoft	1,175	±4	15,881	—
31	—	jamba-1.5-miniOpenAI21 Labs	1,160	±6	9,267	—
32	—	ministral-8b-2410OpenMistral	1,160	±8	5,112	—
33	—	qwen1.5-110b-chatOpenAlibaba	1,155	±4	27,446	—
34	—	c4ai-aya-expanse-8bOpenCohere	1,151	±5	10,394	—
35	—	mixtral-8x22b-instruct-v0.1OpenMistral	1,149	±3	53,760	—
36	—	qwen1.5-72b-chatOpenAlibaba	1,148	±4	40,633	—
37	—	command-rOpenCohere	1,148	±3	56,357	—
38	—	llama-3.1-tulu-3-8bOpenAi2	1,147	±9	3,074	—
39	—	llama-3-8b-instructOpenMeta	1,143	±3	109,216	—
40	—	granite-3.1-8b-instructOpenIBM	1,135	±10	3,299	—
41	—	yi-1.5-34b-chatOpen01 AI	1,133	±4	25,123	—
42	—	llama-3.1-8b-instructOpenMeta	1,133	±3	52,625	—
43	—	zephyr-orpo-141b-A35b-v0.1OpenHuggingFace	1,130	±9	4,858	—
44	—	qwen1.5-32b-chatOpenAlibaba	1,126	±3	22,762	—
45	—	phi-3-medium-4k-instructOpenMicrosoft	1,118	±4	26,089	—
46	—	gemma-2-2b-itOpenGoogle	1,118	±3	48,285	—
47	—	internlm2_5-20b-chatOpenInternLM	1,116	±5	10,594	—
48	—	mixtral-8x7b-instruct-v0.1OpenMistral	1,114	±0	76,092	—
49	—	dbrx-instruct-previewOpenDatabricks	1,113	±3	33,714	—
50	—	granite-3.0-8b-instructOpenIBM	1,110	±7	7,002	—
51	—	qwen1.5-14b-chatOpenAlibaba	1,107	±4	18,665	—
52	—	wizardlm-70bOpenMicrosoft	1,104	±6	8,379	—
53	—	granite-3.1-2b-instructOpenIBM	1,104	±9	3,384	—
54	—	snowflake-arctic-instructOpenSnowflake	1,103	±3	34,177	—
55	—	yi-34b-chatOpen01 AI	1,101	±4	15,928	—
56	—	openchat-3.5-0106OpenOpenChat · 1 more variant	1,100	±5	12,977	—
57	—	gemma-1.1-7b-itOpenGoogle	1,099	±5	25,061	—
58	—	phi-3-small-8k-instructOpenMicrosoft	1,098	±5	18,479	—
59	—	llama-3.2-3b-instructOpenMeta	1,093	±5	8,398	—
60	—	tulu-2-dpo-70bOpenAllenAI/UW	1,092	±7	6,655	—
61	—	starling-lm-7b-betaOpenNexusflow	1,092	±5	16,667	—
62	—	vicuna-33bOpenLMSYS	1,091	±4	22,940	—
63	—	deepseek-llm-67b-chatOpenDeepSeek	1,091	±8	4,986	—
64	—	openhermes-2.5-mistral-7bOpenNousResearch	1,090	±7	5,087	—
65	—	qwq-32b-previewOpenAlibaba	1,088	±10	3,412	—
66	—	starling-lm-7b-alphaOpenUC Berkeley	1,083	±5	10,412	—
67	—	granite-3.0-2b-instructOpenIBM	1,082	±8	7,187	—
68	—	llama-2-70b-chatOpenMeta	1,082	±3	39,614	—
69	—	llama2-70b-steerlm-chatOpenNVIDIA	1,074	±9	3,637	—
70	—	nous-hermes-2-mixtral-8x7b-dpoOpenNousResearch	1,072	±10	3,834	—
71	—	dolphin-2.2.1-mistral-7bOpenCognitive Computations	1,069	±12	1,712	—
72	—	phi-3-mini-4k-instruct-june-2024OpenMicrosoft	1,068	±5	12,810	—
73	—	mistral-7b-instruct-v0.2OpenMistral	1,067	±4	20,053	—
74	—	solar-10.7b-instruct-v1.0OpenUpstage AI	1,066	±8	4,290	—
75	—	qwen1.5-7b-chatOpenAlibaba	1,065	±9	4,863	—
76	—	wizardlm-13bOpenMicrosoft	1,063	±7	7,176	—
77	—	mpt-30b-chatOpenMosaicML	1,059	±13	2,646	—
78	—	falcon-180b-chatOpenTII	1,059	±16	1,328	—
79	—	vicuna-13bOpenLMSYS	1,058	±5	19,778	—
80	—	qwen-14b-chatOpenAlibaba	1,056	±7	5,067	—
81	—	phi-3-mini-4k-instructOpenMicrosoft	1,056	±4	21,087	—
82	—	llama-2-13b-chatOpenMeta	1,054	±5	19,729	—
83	—	smollm2-1.7b-instructOpenHuggingFace	1,051	±13	2,370	—
84	—	zephyr-7b-betaOpenHuggingFace	1,049	±5	11,323	—
85	—	phi-3-mini-128k-instructOpenMicrosoft	1,049	±5	21,620	—
86	—	gemma-7b-itOpenGoogle	1,048	±7	9,179	—
87	—	codellama-34b-instructOpenMeta	1,047	±7	7,506	—
88	—	zephyr-7b-alphaOpenHuggingFace	1,046	±14	1,813	—
89	—	llama-3.2-1b-instructOpenMeta	1,044	±8	8,527	—
90	—	gemma-1.1-2b-itOpenGoogle	1,040	±5	11,351	—
91	—	guanaco-33bOpenUW	1,039	±9	2,995	—
92	—	codellama-70b-instructOpenMeta	1,039	±19	1,192	—
93	—	stripedhyena-nous-7bOpenTogether AI	1,034	±8	5,274	—
94	—	mistral-7b-instructOpenMistral	1,027	±6	9,143	—
95	—	llama-2-7b-chatOpenMeta	1,024	±5	14,551	—
96	—	vicuna-7bOpenLMSYS	1,021	±8	7,014	—
97	—	qwen1.5-4b-chatOpenAlibaba	1,014	±6	7,813	—
98	—	gemma-2b-itOpenGoogle	1,008	±9	4,922	—
99	—	olmo-7b-instructOpenAi2	993	±7	6,507	—
100	—	koala-13bOpenUC Berkeley	979	±8	7,025	—
101	—	chatglm3-6bOpenTsinghua	974	±7	4,760	—
102	—	gpt4all-13b-snoozyOpenNomic AI	972	±13	1,787	—
103	—	mpt-7b-chatOpenMosaicML	967	±9	3,998	—
104	—	alpaca-13bOpenStanford	966	±9	5,866	—
105	—	RWKV-4-Raven-14BOpenRWKV	949	±8	4,919	—
106	—	chatglm2-6bOpenTsinghua	942	±11	2,708	—
107	—	oasst-pythia-12bOpenOpenAssistant	930	±9	6,370	—
108	—	fastchat-t5-3bOpenLMSYS	895	±10	4,287	—
109	—	chatglm-6bOpenTsinghua	888	±9	4,980	—
110	—	dolly-v2-12bOpenDatabricks	881	±10	3,482	—
111	—	llama-13bOpenMeta	867	±13	2,445	—
112	—	stablelm-tuned-alpha-7bOpenStability AI	858	±11	3,334	—

62 more models

Intelligence over time

Each maker's best model at the time on the Artificial Analysis Intelligence Index, by the month it was released.

Gained most · Anthropic +57.2Highest now · Anthropic 64.6Gap between highest and lowest 18.1 ptsTime range Nov 2025 – Nov 2026

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