AlibabaWan 3.0
Alibaba's video model for long clips steered by images, videos and audio, from 4 credits a second.
What is Wan 3.0?
Wan 3.0 is Alibaba's video model, built to combine image, video and audio references into one clip.
On Nsketch it renders 2 to 30 seconds at 480p, 720p or 1080p, with sound. Give it up to 10 reference images, 5 reference videos and 5 audio clips, or set the first and last frames instead. It can also match its length to your reference video.
It's one of the least expensive ways to make long clips on Nsketch, starting at 4 credits a second at 480p. Wan 3.0 Prime, in the same picker, trades a higher price for better detail.
- Provider
- Alibaba
- Length
- 2 to 30 seconds
- Resolution
- 480p, 720p, 1080p
- Aspect ratios
- 16:9, 9:16, 1:1, 4:3, 3:4
- Audio
- On or off, same price
What you can do with Wan 3.0
Images, videos and audio
Up to 10 images, 5 videos and 5 audio clips as references in one generation.
First and last frame
Fix the opening and closing frames and let Wan fill in the motion.
Length that matches your clip
Set the length to auto and it follows your reference video.
Wan 3.0 Prime
A higher-detail version of the same model for final renders.
Prompts to try with Wan 3.0
“The person from the reference photo performs the dance from the reference video on a rooftop at sunset, city skyline behind them, same camera framing as the reference. Music from the reference audio. 15 seconds, 9:16.”
Where Wan 3.0 fits best
Trends and dances
Put a person from a photo into the moves of a reference clip, with the original audio.
Long clips on a budget
Draft 30-second scenes at 480p for a fraction of what other models cost.
Clips shaped by your audio
Add music or a sound clip as a reference to guide the sound of the video.
How Wan 3.0 compares
| Model | Length | Top resolution | Native audio |
|---|---|---|---|
 Seedance 2.5 | 30 seconds | 1080p | |
 Kling 4.0 | 30 seconds | 1080p | |
 Veo 3.1 | 8 seconds | 1080p | |
| Wan 3.0 | 30 seconds | 1080p | |
| Seedance 2.0 | 15 seconds | 4K | |
| Kling 3.0 | 15 seconds | 1080p | |
 Sora 2 | 20 seconds | 720p | |
 Grok Imagine | 15 seconds | 720p | |
| Gemini Omni Flash | 10 seconds | 4K |
Good to know before you generate
- Credits are per second: 4 a second at 480p, 8 at 720p and 14 at 1080p. Wan 3.0 Prime costs 6, 12 and 26.
- Reference video seconds are billed along with the output seconds.
- Reference videos and audio clips are each capped at 15 seconds in total.
- The output and any reference video add up to 30 seconds at most, so a 15-second reference allows up to 15 seconds of output.
- Use either first and last frames or reference media in one generation, not both.
Start creating with Wan 3.0
Create with Wan 3.0Pick Wan 3.0
Open the video generator; Wan 3.0 is already selected when you come from this page.
Describe it
Write what you want to see, and add reference images if you have them.
Generate
You see the credit cost before you generate, and a failed generation gives its credits back.
What people say about Nsketch
Good app
“I tried the latest templates and the quality is really good”
Nice app
“Very nice app! Good quality and nice Trends”
Fun
“Fun and useful to creating moments that my family can’t make”
Great Quality
“The videos that came have really great quality but they took a little long to generate”
noice
“W nsketch ai best ai tool”
Review for this app
“Amazing app”
Plans and pricing
Every plan includes Wan 3.0. Credits work across every model on Nsketch.
Starter
$11/mo
Billed annually at $132
Get Starter- 600 credits/mo
- 1 hour 5 minutes of TTS / voice cloning
- 2 Concurrent generations
- Up to ~75 Image Generations/month
- Up to ~30 Video Generations/month
- Image Editing
Pro
$33/mo
Billed annually at $399
Get Pro- 2,000 credits/mo
- 4 hours 10 minutes of TTS / voice cloning
- 6 Concurrent generations
- Up to ~250 Image Generations/month
- Up to ~100 Video Generations/month
- Image Editing
Max
$67/mo
Billed annually at $799
Get Max- 4,200 credits/mo
- 7 hours 35 minutes of TTS / voice cloning
- 8 Concurrent generations
- Up to ~525 Image Generations/month
- Up to ~210 Video Generations/month
- Image Editing
FAQ
Questions about Wan 3.0
How long can a Wan 3.0 video be?
From 2 to 30 seconds, or set to auto to match the length of your reference video.
What resolutions does Wan 3.0 support?
480p, 720p and 1080p, in 16:9, 9:16, 1:1, 4:3 or 3:4.
What references can I use with Wan 3.0?
Up to 10 images, 5 videos and 5 audio clips, or a first and last frame instead. Reference videos and audio are each limited to 15 seconds in total.
What is Wan 3.0 Prime?
A higher-detail version of Wan 3.0 with the same inputs, priced at 6 to 26 credits a second depending on resolution.
How much does Wan 3.0 cost on Nsketch?
4 credits a second at 480p, 8 at 720p and 14 at 1080p, plus the length of any reference video.
Who makes Wan 3.0?
Alibaba's Wan team.
More video models on Nsketch
Seedance 2.5ByteDance's longest single take: up to 30 seconds with sound, guided by images, clips and audio you upload.
Kling 4.0Kling's newest model: clips up to 30 seconds with sound, steered by your own images and videos.
Veo 3.1Google's video model for short, polished clips with dialogue, sound effects and music.
Seedance 2.0ByteDance's Seedance 2.0 family: Standard up to 4K, plus Fast and Mini for cheaper drafts.