1
O
gpt-image-2 (medium)
gpt-image-2 (medium)Openai
100.0
8.6K
-
¥36.4 / ¥233Input/Output
Image · Text-to-Image · Portraits Leaderboard
Ranking for Text-to-Image / Portraits, based on public preference data.
Portraits model ranking guide
Ranking for Text-to-Image / Portraits, based on public preference data.
gpt-image-2 (medium)grok-imagine-image-qualitygemini-3.1-flash-image-preview (nano-banana-2) [web-search]mai-image-2.5-previewgpt-image-1.5-high-fidelity
2
X
grok-imagine-image-quality
grok-imagine-image-qualityXai
98.4
4.2K
1M
¥9 / ¥18Input/Output
3
G
gemini-3.1-flash-image-preview (nano-banana-2) [web-search]
gemini-3.1-flash-image-preview (nano-banana-2) [web-search]Google
96.8
15.7K
131K
¥3.6 / ¥21.6Input/Output
4
M
mai-image-2.5-preview
mai-image-2.5-previewMicrosoft Ai
95.2
1.1K
32K
¥36 / ¥338Input/Output
5
Ogpt-image-1.5-high-fidelity
gpt-image-1.5-high-fidelityOpenai
93.7
23.4K
1M
¥36 / ¥216Input/Output
6
Ggemini-3-pro-image-preview-2k (nano-banana-pro)
gemini-3-pro-image-preview-2k (nano-banana-pro)Google
92.1
25.5K
1.05M
¥14.4 / ¥86.4Input/Output
7
Ggemini-3-pro-image-preview (nano-banana-pro)
gemini-3-pro-image-preview (nano-banana-pro)Google
90.5
20.7K
1.05M
¥14.4 / ¥86.4Input/Output
8
A
qwen-image-2.0-pro-2026-04-22
qwen-image-2.0-pro-2026-04-22Alibaba
88.9
1.3K
8.19K
¥0 / ¥0Input/Output
9
Mmai-image-2
mai-image-2Microsoft Ai
87.3
10.8K
32K
¥36 / ¥338Input/Output
10
Xgrok-imagine-image
grok-imagine-imageXai
85.7
38.5K
1M
¥9 / ¥18Input/Output
11
L
uni-1.1
uni-1.1Luma Ai
84.1
2.1K
-
-
12
Xgrok-imagine-image-pro
grok-imagine-image-proXai
82.5
21K
1M
¥9 / ¥18Input/Output
13
Ggemini-2.5-flash-image-preview (nano-banana)
gemini-2.5-flash-image-preview (nano-banana)Google
81.0
216.4K
1.05M
¥2.16 / ¥18Input/Output
14
Gimagen-ultra-4.0-generate-001
imagen-ultra-4.0-generate-001Google
79.4
112.2K
-
-
15
R
recraft-v4.1-utility-pro
recraft-v4.1-utility-proRecraft
77.8
510
-
-
16
B
seedream-4.5
seedream-4.5Bytedance
76.2
43.3K
-
-
17
Luni-1.1-max
uni-1.1-maxLuma Ai
74.6
2K
-
-
18
BF
flux-2-max
flux-2-maxBfl
73.0
25.6K
-
-
19
R
reve-v1.5
reve-v1.5Reve
71.4
3K
-
-
20
Aqwen-image-2512
qwen-image-2512Alibaba
69.8
15.7K
8.19K
¥3.6 / ¥14.4Input/Output
21
BFflux-2-pro
flux-2-proBfl
68.3
33.6K
-
-
22
Gimagen-4.0-generate-001
imagen-4.0-generate-001Google
66.7
141.8K
-
-
23
TE
hunyuan-image-3.0
hunyuan-image-3.0Tencent
65.1
42.5K
-
-
24
Rrecraft-v4.1-pro
recraft-v4.1-proRecraft
63.5
595
-
-
25
Bseedream-4-2k
seedream-4-2kBytedance
61.9
3.7K
-
-
26
BFflux-2-flex
flux-2-flexBfl
60.3
32.4K
-
-
27
Bseedream-5.0-lite
seedream-5.0-liteBytedance
58.7
15.2K
-
-
28
Az-image-turbo
z-image-turboAlibaba
57.1
4.3K
-
-
29
Aqwen-image-2.0-2026-03-03
qwen-image-2.0-2026-03-03Alibaba
55.6
12.5K
8.19K
¥0 / ¥0Input/Output
30
Bseedream-4-high-res-fal
seedream-4-high-res-falBytedance
54.0
40.2K
-
-
31
BFflux-2-dev
flux-2-devBfl
52.4
13.3K
-
-
32
Awan2.7-image-pro
wan2.7-image-proAlibaba
50.8
6.8K
8.19K
¥0 / ¥0Input/Output
33
Awan2.7-image
wan2.7-imageAlibaba
49.2
6.9K
8.19K
¥0 / ¥0Input/Output
34
Awan2.6-t2i
wan2.6-t2iAlibaba
47.6
33.5K
-
-
35
Rrecraft-v4
recraft-v4Recraft
46.0
15.8K
-
-
36
Bseedream-4-fal
seedream-4-falBytedance
44.4
3.6K
-
-
37
Ogpt-image-1
gpt-image-1Openai
42.9
65.6K
-
¥36 / ¥288Input/Output
38
Ogpt-image-1-mini
gpt-image-1-miniOpenai
41.3
33.1K
1M
¥36 / ¥216Input/Output
39
HI
hidream-o1-image
hidream-o1-imageHidream
39.7
1.1K
200K
¥108 / ¥432Input/Output
40
Bseedream-3
seedream-3Bytedance
38.1
10.5K
-
-
41
UNp-image
-
36.5
22.2K
-
-
42
Mmai-image-1
mai-image-1Microsoft Ai
34.9
19.5K
32K
¥36 / ¥338Input/Output
43
Awan2.5-t2i-preview
wan2.5-t2i-previewAlibaba
33.3
45.1K
-
-
44
BFflux-2-klein-9b
flux-2-klein-9bBfl
31.7
30.9K
-
-
45
BFflux-1-kontext-max
flux-1-kontext-maxBfl
30.2
19.9K
-
-
46
Gimagen-3.0-generate-002
imagen-3.0-generate-002Google
28.6
89.4K
-
-
47
Aqwen-image
qwen-imageAlibaba
27.0
29K
8.19K
¥3.6 / ¥14.4Input/Output
48
BFflux-1-kontext-pro
flux-1-kontext-proBfl
25.4
96.6K
-
-
49
BFflux-2-klein-4b
flux-2-klein-4bBfl
23.8
31K
-
-
50
I
ideogram-v3-quality
ideogram-v3-qualityIdeogram
22.2
28.4K
-
-
51
Aqwen-image-prompt-extend
qwen-image-prompt-extendAlibaba
20.6
197.5K
8.19K
¥3.6 / ¥14.4Input/Output
52
BFflux-1.1-pro
flux-1.1-proBfl
19.0
14.2K
-
-
53
LA
lucid-origin
lucid-originLeonardo Ai
17.5
74.2K
-
-
54
Rrecraft-v3
recraft-v3Recraft
15.9
43.5K
-
-
55
R
runway-gen4
runway-gen4Runway
14.3
11.2K
-
-
56
Lphoton
photonLuma Ai
12.7
31.6K
-
-
57
Iideogram-v2
ideogram-v2Ideogram
11.1
14.5K
-
-
58
Z
glm-image
glm-imageZai
9.5
1K
200K
¥5.4 / ¥18.7Input/Output
59
BFflux-1-dev-fp8
flux-1-dev-fp8Bfl
7.9
10.6K
-
-
60
Ggemini-2.0-flash-preview-image-generation
gemini-2.0-flash-preview-image-generationGoogle
6.3
76.2K
1.05M
¥1.08 / ¥4.32Input/Output
61
BFflux-1-kontext-dev
flux-1-kontext-devBfl
4.8
64.4K
-
-
62
UNstable-diffusion-v35-large
-
3.2
5.7K
0.26K
¥0.58 / ¥0Input/Output
63
Bbagel
bagelBytedance
1.6
3.5K
-
-
64
Odall-e-3
dall-e-3Openai
0.0
60.9K
-
-
gpt-image-2 (medium) why it ranks first
gpt-image-2 (medium) ranks first with a percent score of 100.0 and 8.6K samples. Use it as the first option for this leaderboard, then compare price, context and availability.
Do not only look at rank #1
Start with the leaderboard closest to your task. Compare the top models by score and sample size, then check price, context length, open or closed access, and provider availability.
Compare adjacent capabilities
FAQ
人像排行榜看什么指标？
主要看排名、百分制分数、样本量和来源。分数用于快速比较同一榜单内模型表现，样本量用于判断结果稳定性。
为什么不同榜单不能直接混合成总分？
不同榜单的任务、样本和评测口径不同，模力榜默认只在同一榜单内排序，避免把写作、代码、图像等能力强行合并。
人像模型应该怎么选？
优先看与你任务最接近的榜单，再结合价格、上下文长度、开源闭源和厂商可用性。排名靠前不代表适合所有预算和部署方式。
榜单多久更新？
页面展示的是最新成功采集的公开榜单数据。当前优先使用 LMArena leaderboard dataset，并在页面来源中保留原始链接。