1
B
dreamina-seedance-2.0-720p
dreamina-seedance-2.0-720pBytedance
100.0
29K
-
-
Video · Text-to-Video · Overall Leaderboard
Ranking for Text-to-Video / Overall, based on public preference data.
Overall model ranking guide
Ranking for Text-to-Video / Overall, based on public preference data.
dreamina-seedance-2.0-720phappyhorse-1.0wan2.7-t2vveo-3.1-audio-1080psora-2-pro
2
A
happyhorse-1.0
happyhorse-1.0Aorizon
97.4
11.7K
-
-
3
A
wan2.7-t2v
wan2.7-t2vAlibaba
94.9
1.2K
-
-
4
G
veo-3.1-audio-1080p
veo-3.1-audio-1080pGoogle
92.3
16.7K
200K
¥14.4 / ¥57.6Input/Output
5
O
sora-2-pro
sora-2-proOpenai
89.7
35.5K
-
-
6
Gveo-3.1-audio
veo-3.1-audioGoogle
87.2
13.7K
200K
¥14.4 / ¥57.6Input/Output
7
Gveo-3.1-fast-audio
veo-3.1-fast-audioGoogle
84.6
39.3K
200K
¥14.4 / ¥57.6Input/Output
8
Gveo-3.1-fast-audio-1080p
veo-3.1-fast-audio-1080pGoogle
82.1
16.8K
200K
¥14.4 / ¥57.6Input/Output
9
X
grok-imagine-video-720p
grok-imagine-video-720pXai
79.5
121.4K
1M
¥9 / ¥18Input/Output
10
Gveo-3-fast-audio
veo-3-fast-audioGoogle
76.9
25.2K
200K
¥14.4 / ¥57.6Input/Output
11
Gveo-3-audio
veo-3-audioGoogle
74.4
19K
200K
¥14.4 / ¥57.6Input/Output
12
Awan2.6-t2v
wan2.6-t2vAlibaba
71.8
27.5K
-
-
13
Osora-2
sora-2Openai
69.2
47.6K
-
-
14
Awan2.5-t2v-preview
wan2.5-t2v-previewAlibaba
66.7
15.8K
-
-
15
Bseedance-v1.5-pro
seedance-v1.5-proBytedance
64.1
63.9K
-
-
16
Gveo-3
veo-3Google
61.5
14.9K
200K
¥14.4 / ¥57.6Input/Output
17
Gveo-3-fast
veo-3-fastGoogle
59.0
15.2K
200K
¥14.4 / ¥57.6Input/Output
18
UNpixverse-v5.6
-
56.4
23.5K
-
-
19
R
runway-gen-4.5
runway-gen-4.5Runway
53.8
23.5K
-
-
20
K
kling-2.5-turbo-1080p
kling-2.5-turbo-1080pKling
51.3
2.1K
200K
¥108 / ¥432Input/Output
21
Kkling-2.6-pro
kling-2.6-proKling
48.7
63K
-
-
22
UNp-video
-
46.2
7K
-
-
23
Kkling-o1-pro
kling-o1-proKling
43.6
1.2K
200K
¥108 / ¥432Input/Output
24
L
ray-3
ray-3Luma Ai
41.0
1.1K
-
-
25
M
hailuo-2.3
hailuo-2.3Minimax
38.5
53.3K
-
-
26
Mhailuo-02-pro
hailuo-02-proMinimax
35.9
9.4K
-
-
27
Bseedance-v1-pro
seedance-v1-proBytedance
33.3
12.1K
-
-
28
Mhailuo-02-standard
hailuo-02-standardMinimax
30.8
9.3K
-
-
29
K
kandinsky-5.0-t2v-pro
kandinsky-5.0-t2v-proKandinsky
28.2
2K
-
-
30
TE
hunyuan-video-1.5
hunyuan-video-1.5Tencent
25.6
4.3K
200K
¥108 / ¥432Input/Output
31
Gveo-2
veo-2Google
23.1
6.5K
-
-
32
Kkling-v2.1-master
kling-v2.1-masterKling
20.5
14K
-
-
33
UNltx-2-19b
-
17.9
45.9K
-
-
34
Awan-v2.2-a14b
wan-v2.2-a14bAlibaba
15.4
10.4K
-
-
35
Kkandinsky-5.0-t2v-lite
kandinsky-5.0-t2v-liteKandinsky
12.8
1.5K
-
-
36
Bseedance-v1-lite
seedance-v1-liteBytedance
10.3
16.2K
-
-
37
Osora
soraOpenai
7.7
4.1K
-
-
38
Lray2
ray2Luma Ai
5.1
5.2K
-
-
39
PI
pika-v2.2
pika-v2.2Pika
2.6
5.7K
-
-
40
G
mochi-v1
mochi-v1Genmo
0.0
5.9K
-
-
dreamina-seedance-2.0-720p why it ranks first
dreamina-seedance-2.0-720p ranks first with a percent score of 100.0 and 29K 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
文生视频总榜排行榜看什么指标？
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。
文生视频总榜模型应该怎么选？
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。