模模力榜

Chat · Vision · Creative Writing Vision Leaderboard

Ranking for Vision / Creative Writing Vision, based on public preference data.

Selection guide

Creative Writing Vision model ranking guide

Ranking for Vision / Creative Writing Vision, based on public preference data.

gemini-3-proclaude-opus-4-6-thinkingmuse-sparkclaude-opus-4-7claude-opus-4-7-thinking
Current DirectoryChat · Vision · Creative Writing Vision
Models65
Published2026/05/18
Arena public preference evaluationOriginal leaderboard: Vision / Creative Writing VisionPublished: 2026/05/18Leaderboard dataset: LMArena latest parquetOpen Arena sourceOpen leaderboard dataset
1
gemini-3-pro
Google
100.0
1K
1.05M
¥14.4 / ¥86.4Input/Output
2
claude-opus-4-6-thinking
Anthropic
98.4
467
1M
¥36 / ¥180Input/Output
3
muse-spark
Meta
96.9
297
-
-
4
claude-opus-4-7
Anthropic
95.3
413
1M
¥36 / ¥180Input/Output
5
claude-opus-4-7-thinking
Anthropic
93.8
440
1M
¥36 / ¥180Input/Output
6
gemini-3.1-pro-preview
Google
92.2
1.1K
1.05M
¥14.4 / ¥86.4Input/Output
7
claude-opus-4-6
Anthropic
90.6
598
1M
¥36 / ¥180Input/Output
8
gemini-3-flash
Google
89.1
1.5K
1.05M
¥3.6 / ¥21.6Input/Output
9
gemini-3-flash (thinking-minimal)
Google
87.5
1.3K
1.05M
¥3.6 / ¥21.6Input/Output
10
grok-4.20-beta-0309-reasoning
Xai
85.9
603
2M
¥14.4 / ¥43.2Input/Output
11
grok-4.20-multi-agent-beta-0309
Xai
84.4
604
2M
¥14.4 / ¥43.2Input/Output
12
gpt-5.5-high
Openai
82.8
254
1.05M
¥36 / ¥216Input/Output
13
gpt-5.4-high
Openai
81.3
317
1.05M
¥18 / ¥108Input/Output
14
claude-sonnet-4-6
Anthropic
79.7
562
1M
¥21.6 / ¥108Input/Output
15
gemma-4-31b
Google
78.1
1.1K
262K
¥3.24 / ¥7.2Input/Output
16
gemini-2.5-pro
Google
76.6
3.4K
1.05M
¥9 / ¥72Input/Output
17
qwen3.5-397b-a17b
Alibaba
75.0
726
262K
¥3.1 / ¥18.6Input/Output
18
gpt-5.5
Openai
73.4
278
1.05M
¥36 / ¥216Input/Output
19
kimi-k2.5-thinking
Moonshot
71.9
857
262K
¥4.32 / ¥21.6Input/Output
20
chatgpt-4o-latest-20250326
Openai
70.3
1.2K
128K
¥18 / ¥72Input/Output
21
kimi-k2.6
Moonshot
68.8
299
262K
¥6.84 / ¥28.8Input/Output
22
gemini-2.5-flash-preview-09-2025
Google
67.2
294
1M
¥2.16 / ¥18Input/Output
23
glm-5v-turbo
Zai
65.6
466
200K
¥0 / ¥0Input/Output
24
gpt-5.4
Openai
64.1
340
1.05M
¥18 / ¥108Input/Output
25
dola-seed-2.0-pro
Bytedance
62.5
430
-
-
26
gpt-5.1
Openai
60.9
696
400K
¥9 / ¥72Input/Output
27
gpt-5.1-high
Openai
59.4
635
400K
¥9 / ¥72Input/Output
28
grok-4.3
Xai
57.8
244
1M
¥9 / ¥18Input/Output
29
gemma-4-26b-a4b
Google
56.3
697
262K
¥0.94 / ¥2.88Input/Output
30
gemini-2.5-flash
Google
54.7
2.6K
1.05M
¥2.16 / ¥18Input/Output
31
kimi-k2.5-instant
Moonshot
53.1
247
262K
¥4.32 / ¥21.6Input/Output
32
claude-opus-4-20250514
Anthropic
51.6
106
200K
¥108 / ¥540Input/Output
33
gemini-3.1-flash-lite-preview
Google
50.0
976
1.05M
¥1.8 / ¥10.8Input/Output
34
mimo-v2.5
Xiaomi
48.4
391
1.05M
¥2.88 / ¥14.4Input/Output
35
gpt-5.2-chat-latest-20260210
Openai
46.9
761
400K
¥12.6 / ¥101Input/Output
36
grok-4-1-fast-reasoning
Xai
45.3
675
2M
¥1.44 / ¥3.6Input/Output
37
grok-4-0709
Xai
43.8
1.3K
256K
¥21.6 / ¥108Input/Output
38
gpt-5.4-mini-high
Openai
42.2
589
400K
¥5.4 / ¥32.4Input/Output
39
ernie-5.0-preview-1220
Baidu
40.6
381
128K
¥7.92 / ¥14.4Input/Output
40
qwen3.5-27b
Alibaba
39.1
579
262K
¥2.16 / ¥17.3Input/Output
41
qwen3-vl-235b-a22b-instruct
Alibaba
37.5
819
128K
¥2.16 / ¥8.64Input/Output
42
qwen3.5-122b-a10b
Alibaba
35.9
632
262K
¥2.88 / ¥23Input/Output
43
gpt-5.2
Openai
34.4
947
400K
¥12.6 / ¥101Input/Output
44
gpt-5-chat
Openai
32.8
1.3K
400K
¥9 / ¥72Input/Output
45
gpt-5.2-high
Openai
31.3
958
400K
¥12.6 / ¥101Input/Output
46
gpt-4.1-2025-04-14
Openai
29.7
1.3K
1.05M
¥14.4 / ¥57.6Input/Output
47
o3-2025-04-16
Openai
28.1
1.6K
200K
¥14.4 / ¥57.6Input/Output
48
mistral-medium-2508
Mistral
26.6
1.5K
262K
¥2.88 / ¥14.4Input/Output
49
gpt-5-high
Openai
25.0
1.4K
400K
¥9 / ¥72Input/Output
50
gemini-2.5-flash-lite-preview-09-2025-no-thinking
Google
23.4
316
1.05M
¥0.72 / ¥2.88Input/Output
51
mistral-small-2506
Mistral
21.9
461
262K
¥2.88 / ¥14.4Input/Output
52
gpt-5-mini-high
Openai
20.3
1K
400K
¥1.8 / ¥14.4Input/Output
53
gpt-4.1-mini-2025-04-14
Openai
18.8
1.2K
1.05M
¥2.88 / ¥11.5Input/Output
54
gemini-2.5-flash-lite-preview-06-17-thinking
Google
17.2
1.2K
65.5K
¥0.72 / ¥2.88Input/Output
55
gemma-3-27b-it
Google
15.6
630
128K
¥2.15 / ¥2.15Input/Output
56
mimo-v2-omni
Xiaomi
14.1
473
262K
¥2.88 / ¥14.4Input/Output
57
gemini-2.0-flash-001
Google
12.5
247
1.05M
¥1.08 / ¥4.32Input/Output
58
o4-mini-2025-04-16
Openai
10.9
1.3K
200K
¥7.92 / ¥31.7Input/Output
59
gpt-5.4-nano-high
Openai
9.4
553
400K
¥1.44 / ¥9Input/Output
60
hunyuan-vision-1.5-thinking
Tencent
7.8
191
-
-
61
mistral-medium-2505
Mistral
6.3
487
262K
¥2.88 / ¥14.4Input/Output
62
glm-4.6v
Zai
4.7
195
128K
¥2.16 / ¥6.48Input/Output
63
llama-4-scout-17b-16e-instruct
Meta
3.1
234
128K
¥1.44 / ¥5.62Input/Output
64
mistral-small-3.1-24b-instruct-2503
Mistral
1.6
782
262K
¥2.88 / ¥14.4Input/Output
65
llama-4-maverick-17b-128e-instruct
Meta
0.0
212
1M
¥1.8 / ¥6.26Input/Output
Top model analysis

gemini-3-pro why it ranks first

gemini-3-pro ranks first with a percent score of 100.0 and 1K samples. Use it as the first option for this leaderboard, then compare price, context and availability.

How to choose

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.

FAQ

FAQ

看图创作排行榜看什么指标？

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。

看图创作模型应该怎么选？

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。