模型排行

排名	變化	模型	得分	置信區間	投票
1	—	gemini-2.5-flash-image-preview (nano-banana)閉源Google	1,145	±3	256,820
2	—	imagen-ultra-4.0-generate-001閉源Google	1,133	±2	248,036
3	—	gpt-image-1閉源OpenAI	1,124	±3	150,073
4	—	imagen-4.0-generate-001閉源Google	1,117	±2	251,756
5	—	qwen-image-prompt-extend開源Alibaba	1,081	±3	213,916
6	—	seedream-3閉源ByteDance	1,076	±3	197,430
7	—	flux-1-kontext-max閉源Black Forest Labs	1,074	±3	78,057
8	—	imagen-3.0-generate-002閉源Google	1,061	±3	290,278
9	—	qwen-image開源Alibaba	1,051	±3	79,388
10	—	flux-1-kontext-pro閉源Black Forest Labs	1,051	±2	205,624
11	—	photon閉源Luma AI	1,034	±3	219,075
12	—	lucid-origin閉源Leonardo.Ai	1,033	±2	141,748
13	—	ideogram-v3-quality閉源Ideogram	1,031	±3	200,505
14	—	recraft-v3閉源Recraft	1,020	±3	290,850
15	—	ideogram-v2閉源Ideogram	1,010	±3	277,638
16	—	flux-1.1-pro閉源Black Forest Labs	1,006	±3	262,313
17	—	gemini-2.0-flash-preview-image-generation閉源Google	987	±3	177,095
18	—	dall-e-3閉源OpenAI	978	±3	253,607
19	—	flux-1-dev-fp8開源Black Forest Labs	968	±3	50,585
20	—	flux-1-kontext-dev開源Black Forest Labs	965	±3	153,384
21	—	stable-diffusion-v35-large開源Stability AI	937	±4	24,050
22	—	bagel開源ByteDance	915	±5	12,351

智力進步

每家廠商當時最強的模型在 Artificial Analysis 智力指數上的分數，按模型釋出的月份畫。

漲得最多 · Anthropic +57.2目前最高 · Anthropic 64.6最高與最低相差 18.1 分時間範圍 2025年11月 – 2026年11月

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