Draft at 480p, finish at 1080p
Lock motion and timing on low-cost 480p drafts, then render the version you keep at 1080p.
12Hrs:
00Min:
00Sec
50% Limited Sale
Upgrade yearlyWan 3.0 AI Video Generator
Turn a prompt or photo into a 2–30 second video at up to 1080p, with dialogue and sound generated in the same pass. Powered by Alibaba's Wan 3.0 model — test draft shots from 14 credits.
14 credits
Explore models
Prompt inspiration
Find your next idea
Browse prompts for video and images. Pick a scene, adapt it to your story, and start creating.
AI generated supertall tower in Manhattan, 430m vertical structure sculpted by computational subtraction process, massive carved-out cavities
Try videoHANDHELD GUERILLA STYLE FOUND FOOTAGE VIDEO. GRITTTY AND NARROW ANGLE. HE IS HOLDING THE IPHONE CAMERA AND IS COMPLETELY OUTSIDE THE FRAME, NO GRAPHIC OVERLAY. FIRST PERSON CAMERA VIEW ONLY
Try videoYoung future astronauts in Space Scout vests and dark heather grey clothing floating and having fun in the zero gravity chambers at Space Camp Huntsville. Excited, joyful, whimsical children's ebook style, inspirational atmosphere.
Try videoUnder the gaze of the full moon, she rises. Stained glass and starlight, where the last colors of sunset still linger in the sky.
Try videoOne garden male gnome reaches and touches the other garden female gnome hand and the sky bursts into the stars and the Milky Way
Try videoA futuristic digital illustration of BlueGoo and Vera, Mother and Daughter Love is Always the Answer
Try videoAdvertise by holding this book with both hands in front. The background is outer space. Starlink satellites are swirling with golden light, and the suit glows in a cyber style. Akinari Tomimasu's logo flashes and becomes cyberized. Finally, it transforms into a cyber android with intense flashing.
Try videoAnimate the exact rocket from the provided image launching from the pad. White rocket with red nose, 'BG' on the side, bright orange flames. Dynamic launch sequence with camera movement.
Try videoReframe the entire scene into a wide landscape horizontal view, expanding and balancing the composition to 16:9 aspect ratio while preserving all key elements and details.
Try videoIridescent cornflowers glowing softly in a dreamy haze… Whispers of gentle light and quiet magic.
Try videoJapanese anime style, cyberpunk full technica, night view of academy city, futuristic giant buildings and neon-lit streetscape, high-tech academy city atmosphere, a silver dragon flying freely in the wind turning into clouds
Try videoRed fire and golden flame. Two blades, one destiny. Victoria and Edmund walk the moonlit shore as the old world burns behind them.
Try videoCapabilities
What You Can Make with Wan 3.0
Ways to start a Wan 3.0 video on Wan3 AI and the controls that shape the result. Everything here is available in the generator today.
Text to Video
Describe the subject, motion, camera and sound. Wan 3.0 renders 2–30 second cinematic takes directly from your prompt, at your chosen aspect ratio.
Image to Video
Upload a portrait or scene as the starting frame. Wan 3.0 animates natural motion while preserving the exact aspect ratio of your image.
First and Last Frame
Anchor both the start and end frames to direct precise camera moves, narrative transitions and dramatic reveals with full control.
Reference Images
Attach up to 9 tagged reference images (Image 1, Image 2) to keep characters, outfits, products and environments consistent across shots.
Native Audio
Synchronized dialogue, ambient sounds and Foley effects generated in the same pass. Put spoken lines in quotes, or switch audio off anytime.
480p to 1080p Output
Test motion with rapid 480p drafts (from 14 credits), then export high-definition 1080p finals in 16:9, 9:16, 1:1, 4:3 or 3:4.
What's new
What Changed in Wan 3.0
Alibaba's Tongyi Lab opened the Wan 3.0 public beta on 6 August 2026 and made it generally available on 24 August 2026. These are the upgrades that matter when you make videos with it.
30 Seconds in One Take
A single generation now runs up to 30 seconds, double the 15-second limit of Wan 2.7. A scene can build, turn and resolve without stitching short clips together.
Sound Generated with the Picture
Wan 3.0 is an audio-visual model: voices, music and effects are generated together with the frames, so they stay in time with the action.
Steadier Characters and Products
Reference images hold faces, clothing, props and layouts more consistently across a long shot, which was the weak spot of earlier long clips.
You Pay per Second
Cost follows length and resolution. A 5-second 480p draft costs 14 credits and the same clip at 1080p costs 55, so you can test ideas cheaply before the final render.
How to use Wan 3.0
Step 01
Choose how to start
Pick text to video, a first frame with an optional last frame, or reference images, then select Wan 3.0 in the model menu.
Step 02
Write a shot brief
Name the subject, what happens over time, one camera move and what we should hear. For 30-second clips, describe a beginning, a turn and an ending.
Step 03
Set length and quality, then generate
Choose 2–30 seconds and 480p, 720p or 1080p. See the exact credit cost before generating, with automatic instant refunds if a task fails.
How to get better Wan 3.0 results
Prompt habits drawn from Alibaba's Wan 3.0 guide and from testing long clips. Each one saves you a regeneration.
Write an arc, not a still
Thirty seconds needs change. Say what happens first, what shifts in the middle and how the shot ends.
One camera move per shot
"Slow dolly in from medium shot to close-up" works better than stacking pans, zooms and orbits in one sentence.
Simple, Transparent Pricing
Credits cover Wan 3.0 and every other model. Subscribe or buy a one-time pack, and see the cost of each video before you generate.
Save 50% with annual billing, plus 30% off every generation
Starter
$29.90
$14.95/moSave 50%
Billed $179.40/year
6,000 credits−30% per generation
For casual creators exploring AI art
- Commercial license included
- 6,000 credits/year
- 30% fewer credits on every generation
- All Wan, Seedance & Seedream models
- Wan 3.0 & Wan 2.6 video
- Seedance 2.0, 2.0 Fast & 1.5 Pro
- Wan 2.7, Seedream 4.5 & 5.0 images
- Credits may not cover heavy use
Popular
Pro
$49.90
$24.95/moSave 50%
Billed $299.40/year
12,000 credits−30% per generation
For creators who generate daily
- Commercial license included
- 12,000 credits/year
- 30% fewer credits on every generation
- All Wan, Seedance & Seedream models
- Wan 3.0 & Wan 2.6 video
- Seedance 2.0, 2.0 Fast & 1.5 Pro
- Wan 2.7, Seedream 4.5 & 5.0 images
- Email support
Premium
$89.90
$44.95/moSave 50%
Billed $539.40/year
22,800 credits−30% per generation
For power users & studios
- Commercial license included
- 22,800 credits/year
- 30% fewer credits on every generation
- All Wan, Seedance & Seedream models
- Wan 3.0 & Wan 2.6 video
- Seedance 2.0, 2.0 Fast & 1.5 Pro
- Wan 2.7, Seedream 4.5 & 5.0 images
- Priority email support
Yearly members pay 30% fewer credits
While a yearly plan is active, every Wan, Seedance and Seedream generation is charged at 70% of the listed credits, rounded up. A 5-second 720p Wan 3.0 video costs 20 credits instead of 28, a 1080p one 39 instead of 55.
Commercial license with every paid plan
Use the videos and images you create in ads, client projects, social media, and products you sell.
FAQ
Frequently Asked Questions
Wan 3.0 is the third-generation video model from Alibaba's Tongyi Lab, the team behind the Wan (Tongyi Wanxiang) series. It entered public beta on 6 August 2026 and became generally available on 24 August 2026 as the API model wan3.0-video. It creates 2–30 second videos at up to 1080p with native audio from text, images and other references.
No. Wan3 AI is an independent platform and is not affiliated with Alibaba. We run Wan 3.0 through a third-party API provider and add a browser-based generator, credit billing and storage for your videos. Alibaba's own channels are wan.video and Alibaba Cloud Model Studio.
Each generation is 2–30 seconds long. You can choose 480p, 720p or 1080p and an aspect ratio of 16:9, 9:16, 1:1, 4:3 or 3:4, or let the model match your first frame. Videos are delivered as MP4 files. Wan 3.0 does not offer 4K output.
Yes. Dialogue, ambient sound and sound effects are generated together with the video. Audio is on by default, and turning it off does not change the price on Wan3 AI.
Wan 3.0 is billed per second of output. On Wan3 AI a 5-second draft costs 14 credits at 480p, 28 at 720p and 55 at 1080p, and a 30-second 1080p clip costs 329 credits. Yearly members pay 30% fewer credits on every generation, for example 20 instead of 28 for 5 seconds at 720p. You always see the exact cost before confirming, and failed tasks are refunded automatically to your balance.
Wan 3.0 is a closed, API-only model: Alibaba has not published its weights, and as of October 2026 the newest open-weight Wan model is Wan 2.2. On Wan3 AI, Wan 3.0 is available on the free plan and limited only by credits: new accounts in eligible regions get 20 sign-up credits, enough for one 5-second Wan 3.0 video at 480p (14 credits), plus daily check-in credits. A plan or credit pack adds more credits.
Text prompts, a first frame with an optional last frame, or up to 9 reference images. Alibaba's model can also read reference video, audio, documents and web pages, but these inputs are not yet available on Wan3 AI.
Yes, with a paid plan or credit pack: any purchase gives you a commercial license for all content in your account. Content created on the free plan is for personal use only. In every case you are responsible for clearing rights to people, brands or music that appear in your videos.
Can't find what you're looking for? Contact our support team
Make Your First Wan 3.0 Video
Write a shot brief or upload a photo, choose your length up to 30s, and generate your scene with native sound. Credits are refunded automatically if a generation fails.