Wan 3.0
One multimodal model for text, image, video, audio, web and document references, with native audio and up to 30 seconds at 1080p.
- Max Duration
- 30 seconds max
- Resolution
- 480p, 720p, 1080p
- Aspect Ratios
- 16:9, 9:16, 1:1, 4:3, 3:4
- Audio
- Native audio
What makes Wan 3.0 unique
Five reference modes in one model
up to 10 images, 5 video clips, 5 audio clips, 1 web link, or 1 document
Native audio
dialogue, effects, and ambience are generated with the video, not layered on after
2 to 30 seconds
any whole duration, priced per second, so short iterations stay cheap
Up to 1080p
480p for drafts, 1080p for final delivery
Five aspect ratios
16:9, 9:16, 1:1, 4:3, and 3:4 from the same model
20,000-character prompts
room for full scene, dialogue, and audio direction
Where it falls short
- –References and start/end frames are mutually exclusive — pick one way to anchor a shot, not both
- –Reference video and audio clips are capped at 15 seconds in total
- –A web link and a document cannot be attached to the same generation
- –1080p costs 4x what 480p costs per second, so drafting at 1080p adds up quickly
Best for
Prompt tips
- 1.
Describe the whole clip as one continuous shot. Say how the scene develops rather than listing separate cuts.
- 2.
Say what you want to hear. Dialogue in quotes, plus ambience, footsteps, or room tone, gets you closer than leaving audio unmentioned.
- 3.
Attach one reference per idea. An image for the character, a clip for the camera move, an audio file for the tone reads more clearly than a pile of similar references.
- 4.
References replace start and end frames, so decide which anchor matters more before you upload.
- 5.
Draft at 480p and re-render the version you like at 1080p — the per-second price quadruples between them.
How Wan 3.0 works
Write a prompt describing the subject, the action, the camera, and the sound you want. Attach references if you have them — images, video clips, audio clips, a web link, or a document — or anchor the shot with a start and end frame instead. Then pick a resolution, an aspect ratio, and any duration from 2 to 30 seconds, and WAN 3.0 renders the clip with audio included.
How to use on Flashloop
Select Wan 3.0
Open Flashloop and choose Wan 3.0 from the model selector in the video creator.
Write your prompt
Describe what you want to see in your video, be as detailed as you like.
Generate & download
Hit generate and your video will be ready in seconds. Download or share directly.
Frequently asked questions
Any whole number of seconds from 2 to 30. It is billed per second, so a 5-second test costs a sixth of a 30-second render at the same resolution.
480p, 720p, and 1080p. 720p is the default on Flashloop; 480p is the cheapest way to iterate and 1080p is the delivery format.
Yes. Audio is native, so dialogue, sound effects, and ambience come out of the same pass as the picture instead of being added later.
Up to 10 reference images, 5 reference video clips, 5 reference audio clips, one web link, or one document. References cannot be combined with start and end frames in the same generation.
Prime is the same model tuned for faster end-to-end generation. It produces the same formats, durations, and reference modes, and costs roughly 1.5x as many credits per second.
Up to 20,000 characters, which is enough to describe a full scene, its dialogue, and its sound design in one go.
Other video models
View all →Veo 3
Generate 8-second videos with native dialogue, sound effects, and ambient audio in one pass.
Kling 3.0
Create up to 15-second multi-shot videos with character consistency, 4K 60fps support, and bilingual audio.
Kling 3.0 Turbo
A faster, more affordable Kling 3.0, text-to-video and image-to-video at 720p or 1080p, up to 15 seconds.
Ready to create with Wan 3.0?
Available on iOS, Android, and web.
Start Creating