Model rankings

Rank	Change	Model	Intelligence	Price $/M	Speed tok/s	First token
1	—	Inkling SmallOpenThinking Machines	27.8	$0.52	184	2.0 s
2	—	Inkling (xhigh)OpenThinking Machines	25.0	$1.8	98	2.4 s
3	—	Nemotron 3 Ultra 550B A55B (Reasoning)OpenNVIDIA	22.9	$1.0	189	2.4 s
4	▲ 2	Gemma 4 31B (Reasoning)OpenGoogle · 1 more variant	19.0	$0.000	35	1.0 s
5	▼ 1	Muse Glimmer (high)OpenMeta	17.5	$0.64	95	1.0 s
6	▼ 1	Gemma 4 26B A4B (Reasoning)OpenGoogle · 1 more variant	16.7	$0.18	—	—
7	—	Granite 4.2 30BOpenIBM	14.8	$0.28	75	0.9 s
8	—	Gemma 4 12B (Reasoning)OpenGoogle · 1 more variant	14.2	$0.15	113	2.4 s
9	—	Apriel-v1.5-15B-ThinkerOpenServiceNow	13.8	$0.000	—	—
10	▲ 2	Apriel-v1.6-15B-ThinkerOpenServiceNow	13.4	$0.000	—	—
11	▼ 1	Nemotron 3.5 LightningOpenNVIDIA	12.9	$0.095	297	0.6 s
12	▼ 1	Nemotron 3 Super 120B A12B (Reasoning)OpenNVIDIA	12.8	$0.31	222	1.7 s
13	▲ 2	Nemotron Cascade 2 30B A3BOpenNVIDIA	11.7	—	—	—
14	▼ 1	gpt-oss-120b (high)OpenOpenAI · 1 more variant	11.6	$0.26	208	0.8 s
15	▼ 1	Granite 4.2 8BOpenIBM	11.1	$0.11	87	0.9 s
16	—	Nemotron 3 Nano Omni 30B A3B ReasoningOpenNVIDIA	10.3	$0.16	—	—
17	▲ 2	Llama 4 MaverickOpenMeta	10.0	$0.42	109	0.8 s
18	▼ 1	gpt-oss-20b (low)OpenOpenAI · 1 more variant	10.0	$0.10	218	0.9 s
19	▼ 1	DiffusionGemma 26B A4BOpenGoogle	9.5	—	—	—
20	—	Granite 4.2 3BOpenIBM	9.1	$0.053	219	0.4 s
21	—	Llama Nemotron Super 49B v1.5 (Reasoning)OpenNVIDIA · 1 more variant	9.0	$0.40	129	4.0 s
22	—	Llama 3.3 Nemotron Super 49B v1 (Reasoning)OpenNVIDIA · 1 more variant	8.9	—	—	—
23	—	Gemma 4 E4B (Reasoning)OpenGoogle · 1 more variant	8.9	$0.040	41	0.8 s
24	—	NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)OpenNVIDIA · 1 more variant	8.9	$0.088	129	1.2 s
25	—	LFM2.5-2.6BOpenLiquid AI	8.4	$0.000	—	—
26	▲ 14	Llama 4 ScoutOpenMeta	8.1	$0.31	100	0.8 s
27	▼ 1	Gemma 4 E2B (Reasoning)OpenGoogle · 1 more variant	7.8	—	—	—
28	▼ 1	Llama 3.3 Instruct 70BOpenMeta	7.7	$0.71	82	1.5 s
29	▼ 1	Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)OpenNVIDIA	7.5	—	—	—
30	▼ 1	NVIDIA Nemotron Nano 12B v2 VL (Reasoning)OpenNVIDIA · 1 more variant	7.5	$0.30	122	3.7 s
31	▼ 1	Granite 4.1 30BOpenIBM	7.4	—	—	—
32	▼ 1	NVIDIA Nemotron Nano 9B V2 (Reasoning)OpenNVIDIA · 1 more variant	7.4	$0.070	117	9.1 s
33	▼ 1	NVIDIA Nemotron 3 Nano 4BOpenNVIDIA	7.4	—	—	—
34	▼ 1	Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)OpenNVIDIA	7.3	—	—	—
35	▼ 1	Llama 3.1 Instruct 405BOpenMeta	7.3	—	—	—
36	▼ 1	LFM2.5-8B-A1BOpenLiquid AI	7.2	$0.000	—	—
37	▼ 1	Llama 3.1 Nemotron Instruct 70BOpenNVIDIA	6.9	$1.2	166	3.3 s
38	▼ 1	Llama 3.1 Instruct 8BOpenMeta	6.9	$0.028	141	0.8 s
39	▼ 1	Llama 3.1 Instruct 70BOpenMeta	6.6	$0.56	66	1.7 s
40	▼ 1	Granite 4.1 8BOpenIBM	6.6	$0.063	66	0.8 s
41	—	R1 1776OpenPerplexity	6.4	—	—	—
42	—	Llama 3.2 Instruct 90B (Vision)OpenMeta	6.4	—	—	—
43	—	Phi-4 Mini InstructOpenMicrosoft	6.3	$0.000	45	0.8 s
44	—	Granite 4.0 H SmallOpenIBM	6.0	$0.11	20	25 s
45	—	LFM2 24B A2BOpenLiquid AI	5.9	—	—	—
46	—	Phi-4OpenMicrosoft	5.9	$0.22	30	2.9 s
47	—	Granite 4.1 3BOpenIBM	5.9	—	—	—
48	—	Phi-3 Mini Instruct 3.8BOpenMicrosoft	5.8	—	—	—
49	—	Gemma 3n E4B Instruct Preview (May '25)OpenGoogle	5.8	—	—	—
50	—	Phi-4 Multimodal InstructOpenMicrosoft	5.8	$0.000	17	0.8 s
51	—	Llama 2 Chat 7BOpenMeta	5.7	$0.10	—	—
52	—	Llama 3.2 Instruct 3BOpenMeta	5.7	—	—	—
53	—	Llama 3 Instruct 70BOpenMeta	5.5	$1.2	—	—
54	—	Arctic InstructOpenSnowflake	5.4	—	—	—
55	—	Llama 3.2 Instruct 11B (Vision)OpenMeta	5.4	$0.35	22	1.6 s
56	—	Llama 2 Chat 70BOpenMeta	5.3	—	—	—
57	—	Llama 2 Chat 13BOpenMeta	5.3	—	—	—
58	—	DBRX InstructOpenDatabricks	5.3	—	—	—
59	—	LFM2.5-1.2B-ThinkingOpenLiquid AI	5.2	—	—	—
60	—	LFM2 2.6BOpenLiquid AI	5.2	—	—	—
61	—	LFM2.5-1.2B-InstructOpenLiquid AI	5.2	—	—	—
62	—	Granite 4.0 H 1BOpenIBM	5.2	—	—	—
63	—	Gemma 3 270MOpenGoogle	5.1	—	—	—
64	—	Granite 4.0 MicroOpenIBM	5.1	—	—	—
65	—	Llama 65BOpenMeta	5.0	—	—	—
66	—	Granite 4.0 1BOpenIBM	5.0	—	—	—
67	—	LFM2 8B A1BOpenLiquid AI	4.9	—	—	—
68	—	Granite 3.3 8B (Non-reasoning)OpenIBM	4.9	$0.085	16	26 s
69	—	Gemma 3 27B InstructOpenGoogle	4.9	$0.000	—	—
70	—	Gemma 3 1B InstructOpenGoogle	4.8	$0.000	—	—
70	—	Gemma 3 4B InstructOpenGoogle	4.8	$0.000	—	—
70	—	Gemma 3n E2B InstructOpenGoogle	4.8	$0.000	—	—
70	—	Gemma 3n E4B InstructOpenGoogle	4.8	—	—	—
70	—	Granite 4.0 350MOpenIBM	4.8	—	—	—
70	—	Granite 4.0 H 350MOpenIBM	4.8	—	—	—
70	—	LFM2 1.2BOpenLiquid AI	4.8	—	—	—
70	—	LFM2.5-VL-1.6BOpenLiquid AI	4.8	$0.000	—	—
70	—	Llama 3 Instruct 8BOpenMeta	4.8	$0.070	—	—
70	—	Llama 3.2 Instruct 1BOpenMeta	4.8	—	—	—
80	—	Gemma 3 12B InstructOpenGoogle	3.8	$0.000	—	—

30 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