Alibaba
Wan 3.0
Best for long cinematic clips with synchronized sound.
Or run a ready prompt
What Wan 3.0 is good at
Wan 3.0 is Alibaba’s current video model. Phosphene runs two versions: Prime for the highest fidelity and Standard at a lower price. Both generate sound with the picture and run much longer than most video models.
Strong at
- Long single clips, well past the usual ten seconds
- Cinematic motion and scenes that hold together
- Sound generated together with the picture
- Scenes guided by reference images and motion clips
Keep in mind
- Long clips at 1080p cost a lot; check the price table first
- Standard costs less but takes longer than Prime
Tips for Wan 3.0
- Describe the whole clip in order: how it opens, what happens, how it ends.
- Try the idea as a short 480p clip first, then render the final longer and sharper.
- Say what the sound should be: a line of dialogue in quotes, music, or the ambience of the place.
Versions in Phosphene
- From 32 credits
Wan 3.0 Prime Video
Alibaba Wan 3.0 Prime text-to-video via Model Studio. Native audio, 480p–1080p, up to 30 seconds.
- From 21 credits
Wan 3.0 Video
Alibaba Wan 3.0 text-to-video via Model Studio. Native audio, 480p–1080p, up to 30 seconds.
Wan 3.0 specs
| Prime | Standard | |
|---|---|---|
| Price per clip | From 32 credits | From 21 credits |
| Length | 5 to 30 seconds | 5 to 30 seconds |
| Resolution | 480p, 720p, 1080p | 480p, 720p, 1080p |
| Aspect ratios | 1:1, 3:4, 4:3, 9:16, 16:9 | 1:1, 3:4, 4:3, 9:16, 16:9 |
| Starts from | Text, a photo, up to 10 reference images or up to 5 reference clips | Text, a photo, up to 10 reference images or up to 5 reference clips |
| Sound | Generated with the video | Generated with the video |
| Typical time | About 20 seconds | About 40 seconds |
Wan 3.0 price per clip
Credits for one clip at each resolution the composer offers, for the shortest, a ten-second and the longest clip.
Wan 3.0 Prime Video
| Resolution | 5 seconds | 10 seconds | 30 seconds |
|---|---|---|---|
| 480p | 32 | 64 | 191 |
| 720p | 64 | 128 | 382 |
| 1080p | 128 | 255 | 764 |
Lengths offered: 5 to 30 seconds. The price grows with length. Starting from a photo or references costs the same as from text.
Wan 3.0 Video
| Resolution | 5 seconds | 10 seconds | 30 seconds |
|---|---|---|---|
| 480p | 21 | 42 | 124 |
| 720p | 42 | 83 | 248 |
| 1080p | 83 | 166 | 496 |
Lengths offered: 5 to 30 seconds. The price grows with length. Starting from a photo or references costs the same as from text.
The composer shows the exact price before you generate.
Questions about Wan 3.0
- How much does Wan 3.0 cost in Phosphene?
- Wan 3.0 Prime Video: from 32 credits for 5 seconds at 480p to 764 for 30 seconds at 1080p. Wan 3.0 Video: from 21 credits for 5 seconds at 480p to 496 for 30 seconds at 1080p. The composer shows the exact price before you generate.
- What is the difference between Wan 3.0 Prime and Standard?
- Prime is the higher-fidelity version and usually finishes sooner. Standard costs less per clip. Both make sound and offer the same lengths and resolutions.
- Can Wan 3.0 make sound?
- Yes. It generates sound together with the picture, so you do not need to add audio afterwards.
- Do I need an Alibaba Cloud account or API key?
- No. Phosphene runs the model for you and charges credits per clip from your Phosphene balance.
Other video models
- Seedance 2.5Coherent long takes with dialogue and synchronized sound
- Kling 3Cinematic camera motion and multi-shot prompts
- MiniMax H3Fast clips with synchronized sound, or 2K with precise camera direction