Model rankings

Rank	Change	Model	Intelligence	Price $/M	Speed tok/s	First token
1	—	o4-mini (high)ClosedOpenAI	69.8	—	—	—
2	—	Gemini 2.5 Pro Preview (Mar' 25)ClosedGoogle	67.8	—	—	—
3	—	o3ClosedOpenAI	66.8	—	—	—
4	—	Grok 3 mini Reasoning (high)ClosedxAI	66.7	—	—	—
5	—	o3-mini (high)ClosedOpenAI · 1 more variant	65.5	—	—	—
6	—	o1ClosedOpenAI	61.9	—	—	—
7	—	Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)OpenNVIDIA	60.8	—	—	—
8	—	Gemini 2.5 Flash Preview (Reasoning)ClosedGoogle	60.5	—	—	—
9	—	DeepSeek R1OpenDeepSeek	60.2	—	—	—
10	—	QwQ 32BOpenAlibaba	58.1	—	—	—
11	—	Claude 3.7 Sonnet (Extended Thinking)ClosedAnthropic · 1 more variant	57.4	—	—	—
12	—	o1-miniClosedOpenAI	53.8	—	—	—
13	—	DeepSeek V3 0324 (Mar' 25)OpenDeepSeek · 1 more variant	53.2	—	—	—
14	—	GPT-4.1 miniClosedOpenAI	52.9	—	—	—
15	—	GPT-4.1ClosedOpenAI	52.6	—	—	—
16	—	Gemini 2.0 Flash Thinking Experimental (Jan '25)ClosedGoogle	52.3	—	—	—
17	—	DeepSeek R1 Distill Qwen 32BOpenDeepSeek	51.9	—	—	—
18	—	Grok 3ClosedxAI	50.6	—	—	—
19	—	Llama 4 MaverickOpenMeta	50.5	—	—	—
20	—	GPT-4o (March 2025, chatgpt-4o-latest)ClosedOpenAI · 2 more variants	50.3	—	—	—
21	—	Gemini 2.0 Pro Experimental (Feb '25)ClosedGoogle	49.2	—	—	—
22	—	DeepSeek R1 Distill Qwen 14BOpenDeepSeek	49.0	—	—	—
23	—	DeepSeek R1 Distill Llama 70BOpenDeepSeek	48.2	—	—	—
24	—	Gemini 2.0 Flash (Feb '25)ClosedGoogle · 1 more variant	48.1	—	—	—
25	—	Reka Flash 3OpenReka AI	47.1	—	—	—
26	—	Qwen2.5 MaxClosedAlibaba	45.3	—	—	—
27	—	Gemini 1.5 Pro (Sep '24)ClosedGoogle · 1 more variant	44.6	—	—	—
28	—	Claude 3.5 Sonnet (Oct '24)ClosedAnthropic	44.4	—	—	—
29	—	SonarClosedPerplexity	43.4	—	—	—
30	—	Llama 4 ScoutOpenMeta	43.0	—	—	—
31	—	Sonar ProClosedPerplexity	42.7	—	—	—
32	—	QwQ 32B-PreviewOpenAlibaba	42.6	—	—	—
33	—	Gemini 2.0 Flash-Lite (Feb '25)ClosedGoogle	41.4	—	—	—
34	—	Llama 3.3 Instruct 70BOpenMeta	41.1	—	—	—
35	—	GPT-4.1 nanoClosedOpenAI	41.0	—	—	—
36	—	Llama 3.1 Instruct 405BOpenMeta	40.5	—	—	—
37	—	Qwen2.5 Instruct 72BOpenAlibaba	40.4	—	—	—
38	—	MiniMax-Text-01OpenMiniMax	40.2	—	—	—
39	—	Phi-4OpenMicrosoft Azure	40.2	—	—	—
40	—	Command AOpenCohere	40.0	—	—	—
41	—	Llama 3.1 Tulu3 405BOpenAllen Institute for AI	39.9	—	—	—
42	—	Llama 3.3 Nemotron Super 49B v1OpenNVIDIA	39.3	—	—	—
43	—	Grok 2 (Dec '24)ClosedxAI	39.2	—	—	—
44	—	Gemini 1.5 Flash (Sep '24)ClosedGoogle · 1 more variant	39.0	—	—	—
45	—	Mistral Large 2 (Nov '24)OpenMistral · 1 more variant	38.3	—	—	—
46	—	Gemma 3 27B InstructOpenGoogle	37.6	—	—	—
47	—	Grok BetaClosedxAI	37.5	—	—	—
48	—	Pixtral LargeOpenMistral	37.4	—	—	—
49	—	Qwen2.5 Instruct 32BOpenAlibaba	37.4	—	—	—
50	—	Llama 3.1 Nemotron Instruct 70BOpenNVIDIA	37.3	—	—	—
51	—	Nova ProClosedAmazon	37.1	—	—	—
52	—	Qwen2.5 Coder Instruct 32BOpenAlibaba	36.3	—	—	—
53	—	GPT-4o miniClosedOpenAI	35.7	—	—	—
54	—	Llama 3.1 Instruct 70BOpenMeta	35.4	—	—	—
55	—	Mistral Small 3.1OpenMistral	35.3	—	—	—
56	—	Mistral Small 3OpenMistral	35.3	—	—	—
57	—	Claude 3 OpusClosedAnthropic	35.1	—	—	—
58	—	Claude 3.5 HaikuClosedAnthropic	34.7	—	—	—
59	—	DeepSeek R1 Distill Llama 8BOpenDeepSeek	34.0	—	—	—
60	—	Gemma 3 12B InstructOpenGoogle	33.8	—	—	—
61	—	Qwen TurboClosedAlibaba	33.6	—	—	—
62	—	Llama 3.2 Instruct 90B (Vision)OpenMeta	33.4	—	—	—
63	—	Qwen2 Instruct 72BOpenAlibaba	32.6	—	—	—
64	—	Nova LiteClosedAmazon	32.5	—	—	—
65	—	Gemini 1.5 Flash-8BClosedGoogle	30.8	—	—	—
66	—	DeepHermes 3 - Mistral 24B PreviewOpenNous Research	30.0	—	—	—
67	—	Jamba 1.5 LargeOpenAI21 Labs	29.3	—	—	—
68	—	Hermes 3 - Llama-3.1 70BOpenNous Research	29.1	—	—	—
69	—	Jamba 1.6 LargeOpenAI21 Labs	28.8	—	—	—
70	—	Nova MicroClosedAmazon	28.3	—	—	—
71	—	Yi-LargeClosed01.AI	27.9	—	—	—
72	—	Claude 3 SonnetClosedAnthropic	27.8	—	—	—
73	—	Codestral (Jan '25)ClosedMistral · 1 more variant	27.6	—	—	—
74	—	Llama 3 Instruct 70BOpenMeta	27.5	—	—	—
75	—	Mistral Small (Sep '24)OpenMistral · 1 more variant	27.4	—	—	—
76	—	Phi-4 Multimodal InstructOpenMicrosoft Azure	26.9	—	—	—
77	—	Qwen2.5 Coder Instruct 7BOpenAlibaba	26.7	—	—	—
78	—	Mistral Large (Feb '24)ClosedMistral	26.4	—	—	—
79	—	Mixtral 8x22B InstructOpenMistral	26.2	—	—	—
80	—	Phi-4 Mini InstructOpenMicrosoft Azure	26.0	—	—	—
81	—	Phi-3 Medium Instruct 14BOpenMicrosoft Azure	24.5	—	—	—
82	—	Gemma 3 4B InstructOpenGoogle	24.5	—	—	—
83	—	Claude 2.1ClosedAnthropic	24.1	—	—	—
84	—	Llama 3.1 Instruct 8BOpenMeta	23.7	—	—	—
85	—	Pixtral 12B (2409)OpenMistral	23.4	—	—	—
86	—	Mistral MediumClosedMistral	22.8	—	—	—
87	—	Ministral 8BOpenMistral	22.3	—	—	—
88	—	Gemma 2 9BOpenGoogle	22.2	—	—	—
89	—	Phi-3 Mini Instruct 3.8BOpenMicrosoft Azure	22.1	—	—	—
90	—	LFM 40BClosedLiquid AI	21.7	—	—	—
91	—	Command-R+ (Aug '24)OpenCohere · 3 more variants	21.5	—	—	—
92	—	Llama 3 Instruct 8BOpenMeta	21.5	—	—	—
93	—	Gemini 1.0 ProClosedGoogle	20.6	—	—	—
94	—	Aya Expanse 32BOpenCohere	20.1	—	—	—
95	—	Llama 2 Chat 13BOpenMeta	19.9	—	—	—
96	—	DBRX InstructOpenDatabricks	19.8	—	—	—
97	—	Ministral 3BClosedMistral	19.6	—	—	—
98	—	Mistral NeMoOpenMistral	19.6	—	—	—
99	—	Llama 3.2 Instruct 3BOpenMeta	19.5	—	—	—
100	—	DeepSeek R1 Distill Qwen 1.5BOpenDeepSeek	19.2	—	—	—
101	—	Jamba 1.5 MiniOpenAI21 Labs	18.4	—	—	—
102	—	Jamba 1.6 MiniOpenAI21 Labs	17.7	—	—	—
103	—	Mixtral 8x7B InstructOpenMistral	17.0	—	—	—
104	—	DeepHermes 3 - Llama-3.1 8B PreviewOpenNous Research	16.2	—	—	—
105	—	Aya Expanse 8BOpenCohere	16.0	—	—	—
106	—	Codestral-MambaOpenMistral	13.9	—	—	—
107	—	Mistral 7B InstructOpenMistral	10.1	—	—	—
108	—	Llama 3.2 Instruct 1BOpenMeta	9.7	—	—	—
109	—	Llama 2 Chat 7BOpenMeta	8.3	—	—	—

59 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