Model rankings

Rank	Change	Model	Intelligence	Price $/M	Speed tok/s	First token
1	—	Claude Opus 5.5 (Max, Default Fallback)ClosedAnthropic	57.6	$8.0	96	653 s
2	—	Claude Sonnet 5.5 (Max, Default Fallback)ClosedAnthropic	56.0	$4.0	134	372 s
3	—	Claude Opus 5.5 (Xhigh, Default Fallback)ClosedAnthropic	56.0	$8.0	82	110 s
4	—	Muse Wonder Full (Max)ClosedMeta	55.8	—	—	—
5	—	Claude Opus 5.5 (High, Default Fallback)ClosedAnthropic	53.6	$8.0	80	30 s
6	—	Claude Fable 5.1 (Max, Default Fallback)ClosedAnthropic	53.4	$20	70	181 s
7	—	Claude Fable 5.1 (Xhigh, Default Fallback)ClosedAnthropic	53.2	$20	63	74 s
8	—	GPT-6 Astra (Max)ClosedOpenAI	52.7	$20	66	278 s
9	—	Gemini 4 Argon (High)ClosedGoogle	52.6	$4.0	81	57 s
10	—	Muse Wonder (Max)ClosedMeta	52.4	$8.0	162	28 s
11	—	GPT-6 Astra (Xhigh)ClosedOpenAI	52.4	$20	62	162 s
12	—	Claude Sonnet 5.5 (Xhigh, Default Fallback)ClosedAnthropic	51.9	$4.0	108	27 s
13	—	GPT-6.1 Sol (Max)ClosedOpenAI	51.8	$4.0	59	236 s
14	—	Claude Opus 5.5 (Medium, Default Fallback)ClosedAnthropic	51.2	$8.0	83	16 s
15	—	Claude Fable 5.1 (High, Default Fallback)ClosedAnthropic	51.2	$20	58	20 s
16	—	GPT-6.1 Sol (Xhigh)ClosedOpenAI	51.0	$4.0	53	107 s
17	—	GPT-6 Astra (High)ClosedOpenAI	50.9	$20	57	41 s
18	—	GPT-6.1 Sol (High)ClosedOpenAI	50.2	$4.0	53	36 s
19	—	GPT-6 Astra (Medium)ClosedOpenAI	49.6	$20	53	5.1 s
20	—	Claude Fable 5.1 (Medium, Default Fallback)ClosedAnthropic	48.9	$20	55	5.2 s
21	▲ 1	GPT-6.1 Sol (Medium)ClosedOpenAI	47.8	$4.0	51	4.0 s
22	▲ 1	Claude Fable 5.1 (Low, Default Fallback)ClosedAnthropic	46.8	$20	57	1.7 s
23	▲ 1	Claude Sonnet 5.5 (High, Default Fallback)ClosedAnthropic	46.8	$4.0	106	10 s
24	▲ 1	GPT-6 Astra (Low)ClosedOpenAI	45.8	$20	53	2.3 s
25	▲ 2	Claude Haiku 5.5 (Max)ClosedAnthropic	43.4	$0.20	234	322 s
26	▲ 2	Claude Opus 5.5 (Low, Default Fallback)ClosedAnthropic	42.3	$8.0	82	6.2 s
27	▲ 2	GPT-6.1 Sol (Low)ClosedOpenAI	42.1	$4.0	52	2.1 s
28	▲ 2	GPT-5.6 Terra (Max)ClosedOpenAI	42.1	$4.5	130	103 s
29	▲ 2	Claude Haiku 5.5 (Xhigh)ClosedAnthropic	41.2	$0.20	186	60 s
30	▲ 2	Gemini 3.8 Flash (High)ClosedGoogle	40.9	$1.5	137	27 s
31	▲ 2	Claude Sonnet 5.5 (Medium, Default Fallback)ClosedAnthropic	40.8	$4.0	106	1.4 s
32	▲ 2	Gemini 3.8 Flash (Medium)ClosedGoogle	39.8	$1.5	—	—
33	▲ 2	Muse Spark Open (Xhigh)OpenMeta	38.9	$2.0	151	0.8 s
34	▲ 2	GPT-6 Luna (Max)ClosedOpenAI	38.1	$0.20	140	78 s
35	▲ 2	GPT-5.6 Terra (Xhigh)ClosedOpenAI	38.0	$4.5	118	15 s
36	▲ 2	Claude Haiku 5.5 (High)ClosedAnthropic	37.8	$0.20	181	25 s
37	▲ 2	Claude Sonnet 5.5 (Low, Default Fallback)ClosedAnthropic	35.9	$4.0	105	1.0 s
38	▲ 2	GPT-6 Luna (Xhigh)ClosedOpenAI	34.6	$0.20	127	20 s
39	▲ 2	Claude Haiku 5.5 (Medium)ClosedAnthropic	34.5	$0.20	179	9.7 s
40	▲ 2	GPT-5.6 Terra (High)ClosedOpenAI	34.2	$4.5	119	2.2 s
41	▲ 2	Gemini 3.8 Flash (Low)ClosedGoogle	33.5	$1.5	—	—
42	▲ 2	GPT-6 Luna (High)ClosedOpenAI	32.9	$0.20	120	11 s
43	▲ 2	GPT-5.3 Codex (Xhigh)ClosedOpenAI	32.5	$4.8	99	55 s
44	▲ 2	GPT-5.6 Terra (Medium)ClosedOpenAI	30.1	$4.5	114	1.5 s
45	▲ 2	GPT-6 Luna (Medium)ClosedOpenAI	29.9	$0.20	92	2.8 s
46	▲ 2	Gemini 3.1 Pro PreviewClosedGoogle	29.7	$4.5	132	21 s
47	▲ 2	Claude Haiku 5.5 (Low)ClosedAnthropic	29.5	$0.20	173	7.5 s
48	▲ 2	GPT-5.6 Terra (Low)ClosedOpenAI	27.5	$4.5	116	1.3 s
49	▲ 2	Nemotron 3.5 Super VLOpenNVIDIA	26.5	$0.31	446	0.7 s
50	▲ 2	GPT-5.5 Instant (June 2026)ClosedOpenAI	26.0	$11	152	0.9 s
51	▲ 2	Inkling SmallOpenThinking Machines	25.7	$0.52	223	1.6 s
52	▲ 2	Inkling (Xhigh)OpenThinking Machines	25.0	$1.8	195	1.8 s
53	▲ 2	Nemotron 3 Ultra 550B A55B (Reasoning)OpenNVIDIA	22.9	$1.1	158	1.2 s
54	▲ 2	Gemini 3.5 Flash-LiteClosedGoogle	22.2	$0.85	414	7.8 s
55	▲ 2	GPT-6 Luna (Low)ClosedOpenAI	21.5	$0.20	116	2.4 s
56	▲ 2	Nemotron 3.5 Bolt 74B-A9BOpenNVIDIA	21.4	—	—	—
57	▲ 2	GPT-5.6 Terra (Non-reasoning)ClosedOpenAI	20.8	$4.5	122	0.8 s
58	▲ 2	o3ClosedOpenAI	20.2	$3.5	152	7.2 s
59	▲ 2	GPT-6 Luna (Non-reasoning)ClosedOpenAI	18.5	$0.20	120	0.9 s
60	▲ 2	Muse Glimmer (High)OpenMeta	17.5	$0.52	105	1.1 s
61	▲ 2	Gemma 4 26B A4B (Reasoning)OpenGoogle	16.7	$0.16	91	0.9 s
62	▲ 2	Gemma 4 31B (Reasoning)OpenGoogle	14.7	$0.000	35	1.1 s
63	▲ 2	Gemma 4 12B (Reasoning)OpenGoogle	14.2	$0.15	114	2.2 s
64	▲ 2	Nova 2.0 Pro Preview (Medium)ClosedAmazon	14.2	$3.4	127	13 s
65	▲ 2	Gemma 4 31B (Non-reasoning)OpenGoogle	13.9	$0.22	55	1.8 s
66	▲ 2	Nova 2.0 Omni (Medium)ClosedAmazon	13.6	—	—	—
67	▲ 2	Apriel-v1.6-15B-ThinkerOpenServiceNow	13.4	—	—	—
68	▲ 2	Nova 2.0 Lite (High)ClosedAmazon	13.4	$0.85	186	17 s
69	▲ 2	Gemma 4 26B A4B (Non-reasoning)OpenGoogle	13.1	$0.19	108	0.9 s
70	▲ 2	Nemotron 3.5 LightningOpenNVIDIA	12.9	$0.095	295	0.6 s
71	▲ 2	Nova 2.0 Pro Preview (Low)ClosedAmazon	12.8	$3.4	120	11 s
72	▲ 2	Granite 4.2 30BOpenIBM	12.7	$0.28	47	0.8 s
73	▲ 2	Nova 2.0 Lite (Medium)ClosedAmazon	12.5	$0.85	180	13 s
74	▲ 2	Mercury 2.5ClosedInception	12.3	$0.38	714	3.6 s
75	▲ 2	Nova 2.0 Lite (Low)ClosedAmazon	11.8	$0.85	173	9.2 s
76	▲ 2	Nemotron Cascade 2 30B A3BOpenNVIDIA	11.7	—	—	—
77	▲ 2	gpt-oss-120b (High)OpenOpenAI	11.6	$0.26	233	0.8 s
78	▲ 2	Nova 2.0 Omni (Low)ClosedAmazon	11.1	—	—	—
79	▲ 2	Granite 4.2 8BOpenIBM	11.1	$0.11	54	0.8 s
80	▲ 2	Nemotron 3 Nano Omni 30B A3B ReasoningOpenNVIDIA	10.3	$0.64	231	0.8 s
81	▲ 2	gpt-oss-120b (Low)OpenOpenAI	10.2	$0.24	265	0.8 s
82	▲ 2	Nova 2.0 Pro Preview (Non-reasoning)ClosedAmazon	10.0	$3.4	115	1.0 s
83	▲ 2	gpt-oss-20b (Low)OpenOpenAI	10.0	$0.13	186	1.0 s
84	▲ 2	DiffusionGemma 26B A4BOpenGoogle	9.5	—	—	—
85	▲ 2	Gemma 4 12B (Non-reasoning)OpenGoogle	9.4	$0.15	118	2.3 s
86	▲ 2	Llama 4 MaverickOpenMeta	9.3	$0.42	52	1.0 s
87	▲ 2	Nova PremierClosedAmazon	9.2	—	—	—
88	▲ 2	Granite 4.2 3BOpenIBM	9.1	$0.053	51	0.7 s
89	▲ 2	gpt-oss-20b (High)OpenOpenAI	9.0	$0.093	173	0.8 s
90	▲ 2	Gemma 4 E4B (Reasoning)OpenGoogle	8.9	$0.040	45	0.9 s
91	▲ 2	NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)OpenNVIDIA	8.9	$0.088	241	0.9 s
92	▲ 2	Nova 2.0 Lite (Non-reasoning)ClosedAmazon	8.7	$0.85	183	1.0 s
93	▲ 2	LFM2.5-2.6BOpenLiquid AI	8.4	—	—	—
94	▲ 2	Nova 2.0 Omni (Non-reasoning)ClosedAmazon	8.2	—	—	—
95	▲ 2	Gemma 4 E2B (Reasoning)OpenGoogle	7.8	—	—	—
96	▲ 2	Llama 3.3 Instruct 70BOpenMeta	7.7	$0.67	89	1.8 s
97	▲ 2	Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)OpenNVIDIA	7.5	—	—	—
98	▲ 2	NVIDIA Nemotron Nano 12B v2 VL (Reasoning)OpenNVIDIA	7.5	—	—	—
99	▲ 2	Gemma 4 E4B (Non-reasoning)OpenGoogle	7.5	$0.040	45	0.8 s
100	▲ 2	NVIDIA Nemotron Nano 9B V2 (Reasoning)OpenNVIDIA	7.4	$0.070	122	0.6 s
101	▲ 2	Llama 3.1 Instruct 405BOpenMeta	7.3	—	—	—
102	▲ 2	LFM2.5-8B-A1BOpenLiquid AI	7.2	—	—	—
103	▲ 2	NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)OpenNVIDIA	6.8	$0.088	277	0.8 s
104	▲ 2	NVIDIA Nemotron Nano 9B V2 (Non-reasoning)OpenNVIDIA	6.8	$0.086	168	1.0 s
105	▲ 2	Gemma 4 E2B (Non-reasoning)OpenGoogle	6.5	—	—	—
106	▲ 2	Llama 4 ScoutOpenMeta	6.5	$0.30	118	0.8 s
107	▲ 2	R1 1776OpenPerplexity	6.4	—	—	—
108	▲ 2	Llama 3.2 Instruct 90B (Vision)OpenMeta	6.4	—	—	—
109	▲ 2	Phi-4 Mini InstructOpenMicrosoft	6.3	$0.13	46	0.8 s
110	▲ 2	Granite 4.0 H SmallOpenIBM	6.0	—	—	—
111	▲ 2	LFM2 24B A2BOpenLiquid AI	5.9	—	—	—
112	▲ 2	Phi-4OpenMicrosoft	5.9	$0.22	39	2.5 s
113	▲ 2	Nova MicroClosedAmazon	5.9	$0.061	290	0.8 s
114	▲ 2	NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)OpenNVIDIA	5.8	$0.30	202	1.0 s
115	▲ 2	Phi-4 Multimodal InstructOpenMicrosoft	5.8	$0.14	—	—
116	▲ 2	Llama 3.2 Instruct 11B (Vision)OpenMeta	5.4	$0.35	26	1.3 s
117	▲ 2	LFM2.5-1.2B-ThinkingOpenLiquid AI	5.2	—	—	—
118	▲ 2	LFM2.5-1.2B-InstructOpenLiquid AI	5.2	—	—	—
119	▲ 2	Granite 4.0 H 1BOpenIBM	5.2	—	—	—
120	▲ 2	Gemma 3 270MOpenGoogle	5.1	—	—	—
121	▲ 2	Granite 4.0 MicroOpenIBM	5.1	—	—	—
122	▲ 2	Granite 4.0 350MOpenIBM	4.8	—	—	—
122	▲ 2	Granite 4.0 H 350MOpenIBM	4.8	—	—	—
122	▲ 2	LFM2.5-VL-1.6BOpenLiquid AI	4.8	—	—	—
125	▲ 2	NVIDIA Nemotron 3 Nano 4BOpenNVIDIA	4.8	—	—	—

75 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