模型排行

排名	变化	模型	得分	置信区间
1	—	GPT Image 2 (high)闭源OpenAI	1,336	±8
2	—	Reve 2.0闭源Reve	1,278	±11
3	—	MAI-Image-2.5闭源Microsoft AI	1,266	±10
4	—	HiDream-O1-Image-1.5闭源HiDream	1,263	±9
5	—	Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)闭源Google	1,257	±11
6	—	GPT Image 1.5 (high)闭源OpenAI	1,257	±8
7	—	Nano Banana 2 (Gemini 3.1 Flash Image Preview)闭源Google	1,252	±8
8	—	Riverflow 2.0闭源Sourceful	1,249	±9
9	—	Nano Banana Pro (Gemini 3 Pro Image)闭源Google	1,216	±8
10	—	MAI-Image-2.5-Flash闭源Microsoft AI	1,206	±10
11	—	Recraft V4.1 Utility Pro闭源Recraft	1,205	±9
12	—	grok-imagine-image-quality闭源SpaceXAI	1,200	±8
13	—	Krea 2 Medium闭源Krea	1,196	±9
14	—	Recraft V4.1 Utility闭源Recraft	1,195	±9
15	—	FLUX.2 [max]闭源Black Forest Labs	1,191	±8
16	▲ 2	MAI-Image-2闭源Microsoft AI	1,187	±8
17	—	Krea 2 Large闭源Krea	1,186	±9
18	▼ 2	Seedream 4.0闭源ByteDance Seed	1,185	±7
19	—	FLUX.2 [pro]闭源Black Forest Labs	1,185	±8
20	—	FLUX.2 [flex]闭源Black Forest Labs	1,177	±8
21	—	Imagen 4 Ultra Preview 0606闭源Google	1,177	±8
22	—	grok-imagine-image闭源SpaceXAI	1,174	±8
23	—	Luma UNI 1 Max闭源Luma Labs	1,172	±9
24	—	Qwen Image 2.0 Pro (2026-04-22)闭源Alibaba	1,170	±9
25	—	Imagen 4 Ultra闭源Google	1,169	±8
26	—	ImagineArt 2.0闭源ImagineArt	1,167	±8
27	—	Seedream 4.5闭源ByteDance Seed	1,163	±8
28	—	Imagen 4 Preview 0606闭源Google	1,159	±8
29	—	MAI-Image-2-Efficient闭源Microsoft AI	1,155	±8
30	▲ 1	Recraft V4.1 Pro闭源Recraft	1,153	±9
31	▼ 1	Nano Banana (Gemini 2.5 Flash Image)闭源Google	1,153	±8
32	—	Recraft V4.1闭源Recraft	1,148	±9
33	—	ImagineArt 1.5 Preview闭源ImagineArt	1,147	±8
34	—	Luma UNI 1闭源Luma Labs	1,145	±9
35	—	grok-imagine-image-pro闭源SpaceXAI	1,144	±8
36	—	Seedream 3.0闭源ByteDance Seed	1,143	±8
37	—	Wan 2.5 Preview闭源Alibaba	1,142	±8
38	—	Recraft V4 Pro闭源Recraft	1,133	±8
39	—	Wan2.6 Text to Image闭源Alibaba	1,132	±8
40	—	Wan 2.6 Image闭源Alibaba	1,132	±9
41	▲ 1	Recraft V4闭源Recraft	1,131	±8
42	▼ 1	Vivago 2.1闭源HiDream	1,131	±8
43	—	GPT Image 1 (high)闭源OpenAI	1,128	±8
44	—	Kolors 2.1闭源KlingAI	1,126	±8
45	▲ 1	image-1闭源Api Airforce	1,120	±8
46	▼ 1	FLUX.1 Kontext [max]闭源Black Forest Labs	1,119	±8
47	—	Lucid Origin Ultra闭源Leonardo.Ai	1,118	±8
48	—	Seedream 5.0 Lite闭源ByteDance	1,117	±8
49	—	Wan 2.7 Pro闭源Alibaba	1,116	±8
50	—	Lucid Origin Fast闭源Leonardo.Ai	1,109	±8
51	—	Eigen Image闭源Eigen AI	1,105	±8
52	—	Imagen 3 (v002)闭源Google	1,105	±8
53	▲ 1	Wan 2.7闭源Alibaba	1,104	±8
54	▼ 1	Vivago 2.0闭源HiDream	1,102	±8
55	—	Vidu Q2闭源Vidu	1,102	±8
56	—	Qwen Image 2.0 (2026-03-03)闭源Alibaba	1,101	±9
57	—	Imagen 4 Standard闭源Google	1,097	±8
58	—	Reve Image (Halfmoon)闭源Reve	1,094	±8
59	▲ 1	FLUX1.1 [pro] Ultra闭源Black Forest Labs	1,093	±8
60	▼ 1	Kling Image 3.0 Omni闭源KlingAI	1,093	±8
61	—	FLUX.1 Kontext [pro]闭源Black Forest Labs	1,090	±8
62	—	GPT Image 1 Mini (medium)闭源OpenAI	1,086	±8
63	—	Dreamina 3.1闭源ByteDance	1,084	±8
64	—	FLUX1.1 [pro]闭源Black Forest Labs	1,084	±8
65	—	Imagen 4 Fast Preview 0606闭源Google	1,078	±8
66	—	Qwen Image Plus 2601闭源Alibaba	1,076	±8
67	—	Ideogram 3.0闭源Ideogram	1,074	±8
68	—	Imagen 4 Fast闭源Google	1,071	±8
69	—	P-Image闭源Pruna AI	1,068	±8
70	▲ 1	Midjourney v7 Alpha闭源Midjourney	1,065	±8
71	▼ 1	Ideogram v2闭源Ideogram	1,065	±8
72	▲ 3	Recraft V3闭源Recraft	1,065	±8
73	▼ 1	FLUX.1 [pro]闭源Black Forest Labs	1,064	±8
74	—	Midjourney v6闭源Midjourney	1,064	±8
75	▼ 2	Luma Photon闭源Luma Labs	1,064	±8
76	—	Ideogram v2 Turbo闭源Ideogram	1,052	±8
77	—	Image-01闭源MiniMax	1,052	±8
78	—	Midjourney v6.1闭源Midjourney	1,048	±8
79	—	MAI Image 1闭源Microsoft AI	1,039	±8
80	—	Z-Image Base闭源Alibaba	1,031	±8
81	▲ 1	Ideogram v1闭源Ideogram	1,029	±8
82	▼ 1	Phoenix 1.0 Ultra闭源Leonardo.Ai	1,028	±8
83	—	Firefly Image 5 Preview闭源Adobe	1,021	±8
84	—	Stable Diffusion 3 Large闭源Stability.ai	1,020	±8
85	▲ 1	Ideogram v2a Turbo闭源Ideogram	1,012	±9
86	▼ 1	Ideogram v2a闭源Ideogram	1,011	±9
87	—	Playground v3 (beta)闭源Playground AI	1,007	±8
88	—	Phoenix 0.9 Ultra闭源Leonardo.Ai	1,007	±8
89	—	Gemini 2.0 Flash Preview闭源Google	1,004	±12
90	—	Krea 1闭源Krea	1,003	±8
91	—	Firefly Image 4闭源Adobe	1,000	±8
92	—	Phoenix 1.0 Fast闭源Leonardo.Ai	991	±8
93	—	Luma Photon Flash闭源Luma Labs	982	±8
94	▲ 1	Recraft 20B闭源Recraft	975	±8
95	▼ 1	Step Image Edit 2闭源StepFun	975	±11
96	—	Runway Gen-4 Image闭源Runway	974	±9
97	—	DALLE 3 HD闭源OpenAI	956	±8
98	—	Gemini 2.0 Flash Experimental闭源Google	956	±12
99	—	DALLE 3闭源OpenAI	946	±8
100	—	Firefly Image 3闭源Adobe	930	±8
101	—	Grok 2闭源SpaceXAI	920	±14
102	—	Amazon Titan G1 v2 (Standard)闭源Amazon	907	±8
103	—	Stable Diffusion 3 Large Turbo闭源Stability.ai	904	±8
104	—	Stable Diffusion 1.6闭源Stability.ai	899	±9
105	—	Amazon Titan G1 (Standard)闭源Amazon	897	±9
106	—	DALLE 2闭源OpenAI	740	±11

其余 56 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