模型排行

排名	变化	模型	得分	置信区间
1	—	GPT Image 1.5 (high)闭源OpenAI	1,265	±11
2	—	Nano Banana 2 (Gemini 3.1 Flash Image Preview)闭源Google	1,258	±9
3	—	Riverflow 2.0闭源Sourceful	1,254	±11
4	—	Nano Banana Pro (Gemini 3 Pro Image)闭源Google	1,215	±11
5	—	FLUX.2 [max]闭源Black Forest Labs	1,200	±11
6	—	Seedream 4.0闭源ByteDance Seed	1,185	±7
7	—	FLUX.2 [pro]闭源Black Forest Labs	1,182	±12
8	—	FLUX.2 [flex]闭源Black Forest Labs	1,180	±11
9	—	Imagen 4 Ultra Preview 0606闭源Google	1,174	±11
10	▲ 1	grok-imagine-image闭源xAI	1,172	±8
11	▼ 1	Seedream 4.5闭源ByteDance Seed	1,171	±9
12	—	Imagen 4 Preview 0606闭源Google	1,170	±12
13	—	Imagen 4 Ultra闭源Google	1,170	±11
14	—	Nano Banana (Gemini 2.5 Flash Image)闭源Google	1,163	±7
15	—	Seedream 3.0闭源ByteDance Seed	1,149	±10
16	—	Wan 2.5 Preview闭源Alibaba	1,148	±12
17	—	ImagineArt 1.5 Preview闭源ImagineArt	1,148	±8
18	—	grok-imagine-image-pro闭源xAI	1,138	±8
19	—	Wan 2.6 Image闭源Alibaba	1,133	±16
20	—	Wan2.6 Text to Image闭源Alibaba	1,133	±9
21	—	Recraft V4 Pro闭源Recraft	1,132	±9
22	—	GPT Image 1 (high)闭源OpenAI	1,128	±11
23	—	Vivago 2.1闭源HiDream	1,126	±11
24	—	Kolors 2.1闭源KlingAI	1,124	±11
25	—	Recraft V4闭源Recraft	1,121	±9
26	—	Seedream 5.0 Lite闭源ByteDance	1,119	±10
27	—	FLUX.1 Kontext [max]闭源Black Forest Labs	1,115	±9
28	—	Vidu Q2闭源Vidu	1,113	±9
29	—	Lucid Origin Ultra闭源Leonardo.Ai	1,111	±11
30	—	Eigen Image闭源Eigen AI	1,107	±8
31	—	Lucid Origin Fast闭源Leonardo.Ai	1,106	±11
32	—	Vivago 2.0闭源HiDream	1,103	±12
33	—	Imagen 3 (v002)闭源Google	1,103	±10
34	—	Imagen 4 Standard闭源Google	1,097	±11
35	—	Reve Image (Halfmoon)闭源Reve	1,090	±13
36	—	FLUX1.1 [pro] Ultra闭源Black Forest Labs	1,088	±12
37	—	Imagen 4 Fast Preview 0606闭源Google	1,082	±11
38	—	FLUX.1 Kontext [pro]闭源Black Forest Labs	1,082	±12
39	—	Dreamina 3.1闭源ByteDance	1,080	±11
40	—	Ideogram 3.0闭源Ideogram	1,076	±13
41	—	FLUX1.1 [pro]闭源Black Forest Labs	1,076	±9
42	—	Recraft V3闭源Recraft	1,073	±9
43	—	GPT Image 1 Mini (medium)闭源OpenAI	1,071	±11
44	—	FLUX.1 [pro]闭源Black Forest Labs	1,068	±8
45	▲ 2	P-Image闭源Pruna AI	1,067	±13
46	▼ 1	Midjourney v7 Alpha闭源Midjourney	1,066	±13
47	▼ 1	Qwen Image Plus 2601闭源Alibaba	1,066	±8
48	—	Imagen 4 Fast闭源Google	1,065	±11
49	—	Luma Photon闭源Luma Labs	1,064	±10
50	—	Ideogram v2闭源Ideogram	1,057	±8
51	—	Midjourney v6闭源Midjourney	1,047	±6
52	—	Image-01闭源MiniMax	1,046	±13
53	—	Ideogram v2 Turbo闭源Ideogram	1,043	±8
54	—	Midjourney v6.1闭源Midjourney	1,041	±7
55	—	Z-Image Base闭源Alibaba	1,040	±9
56	—	MAI Image 1闭源Microsoft Azure	1,030	±9
57	—	Firefly Image 5 Preview闭源Adobe	1,027	±11
58	—	Stable Diffusion 3 Large闭源Stability.ai	1,024	±6
59	▲ 1	Phoenix 1.0 Ultra闭源Leonardo.Ai	1,020	±12
60	▼ 1	Ideogram v1闭源Ideogram	1,020	±8
61	—	Ideogram v2a闭源Ideogram	1,013	±13
62	—	Phoenix 0.9 Ultra闭源Leonardo.Ai	1,012	±10
63	—	Ideogram v2a Turbo闭源Ideogram	1,008	±13
64	—	Gemini 2.0 Flash Preview闭源Google	1,004	±14
65	—	Playground v3 (beta)闭源Playground AI	1,004	±8
66	—	Krea 1闭源Krea	1,001	±11
67	—	Firefly Image 4闭源Adobe	999	±11
68	▲ 1	Phoenix 1.0 Fast闭源Leonardo.Ai	984	±12
69	▼ 1	Recraft 20B闭源Recraft	984	±8
70	—	Luma Photon Flash闭源Luma Labs	971	±9
71	—	Runway Gen-4 Image闭源Runway	971	±12
72	—	DALLE 3 HD闭源OpenAI	953	±6
73	—	Gemini 2.0 Flash Experimental闭源Google	947	±20
74	—	DALLE 3闭源OpenAI	938	±6
75	—	Firefly Image 3闭源Adobe	935	±7
76	—	Grok 2闭源xAI	917	±25
77	—	Stable Diffusion 3 Large Turbo闭源Stability.ai	909	±6
78	—	Amazon Titan G1 v2 (Standard)闭源Amazon	909	±7
79	—	Stable Diffusion 1.6闭源Stability.ai	902	±6
80	—	Amazon Titan G1 (Standard)闭源Amazon	891	±7
81	—	DALLE 2闭源OpenAI	719	±7

其余 31 个模型

智力进步

每家厂商当时最强的模型在 Artificial Analysis 智力指数上的分数，按模型发布的月份画。

涨得最多 · Anthropic +57.2目前最高 · Anthropic 64.6最高与最低相差 18.1 分时间范围 2025年11月 – 2026年11月

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