Alibaba · Video model
Wan 2.7.
Wan 2.7 is a video generation model from Alibaba. It supports text-to-video, image-to-video with first and last frame control, and reference-to-video, where multiple reference images guide the style and content...
Balanced quality
78
#5 among comparable OpenRouter models
Text to video
1109 Elo
Arena rank #8
Image to video
1092 Elo
Arena rank #5
Price
$9
Per minute at the evaluated setting
What this means.
This model has direct preference evidence for both creating video from text and animating an existing image. The balanced score gives both workflows equal importance.
Practical specifications.
- Resolutions
- 720p, 1080p
- Aspect ratios
- 16:9, 9:16, 1:1, 4:3, 3:4
- Durations
- 2–10 seconds
- Generates audio
- Yes
- First/last frame control
- first_frame, last_frame
See where Wan 2.7 stands.
| Model | Company | Quality | Evidence |
|---|---|---|---|
| Seedance 2.0 | ByteDance | 98 | Both workflows / exact match |
| HappyHorse 1.1 | Alibaba | 88 | Both workflows / exact match |
| HappyHorse 1.0 | Alibaba | 82 | Both workflows / exact match |
| Grok Imagine Video 1.5 | xAI | 80 | Limited |
| Wan 2.7 | Alibaba | 78 | Both workflows / exact match |
| Video v3.0 Pro | Kling AI | 70 | Both workflows / exact match |
| Veo 3.1 | 68 | Both workflows / exact match | |
| Video v3.0 Standard | Kling AI | 60 | Both workflows / exact match |
| Veo 3.1 Fast | 58 | Both workflows / exact match | |
| Grok Imagine Video | xAI | 52 | Both workflows / exact match |
| Veo 3.1 Lite | 46 | Both workflows / exact match | |
| Seedance 1.5 Pro | ByteDance | 28 | Both workflows / exact match |