AlibabaVideo model

Wan 3.0

Images, clips and sound in.

Illustration
Up to 30sClip referencesFrom 42 credits
From 42 credits per 5 secondsGenerate with Wan 3.0

Why Wan 3.0

Every kind of reference

Images for faces and products, clips for motion and camera, audio for voice and music. All addressable from the prompt, all in one generation.

Made for

  • Reference-led ads

    Your product, your presenter and a clip for the camera move, combined into one shot.

  • Recasting footage

    Keep the scene and the moves, change who is in it.

  • Long drafts

    Thirty seconds at 480p for a few hundred credits.

  • Music-driven clips

    Give it the track as @Audio1 and let the picture follow.

The specs

Provider
Alibaba
Resolution
480p, 720p, 1080p
Duration
2 to 30 seconds
Aspect ratios
Auto, 16:9, 9:16, 1:1, 4:3, 3:4
Reference images
Up to 9, as @img1 to @img9
Reference clips
Up to 5, 15 seconds combined, at least 16 fps
Reference audio
Up to 5 tracks, 15 seconds combined
Audio
Native, on or off, same price
Speed
Standard, or Prime for a faster render
Prices
42, 84 or 167 per 5s at 480p, 720p, 1080p
Plans
Starter and up
Cost
From 42 credits per 5 seconds

Straight answers

What is Wan 3.0?

Wan 3.0 is Alibaba's newest Wan video model, and it ranks at the top of public video edit leaderboards. It generates 2 to 30 second clips at 480p, 720p or 1080p from a prompt plus references: up to nine images, five video clips and five audio tracks, called from the prompt as @img1, @Video1 and @Audio1. That makes one model do three jobs: a new shot built from references, an edit of an existing clip, or a recast that puts a new lead into footage. Sound is generated with the picture at no extra cost. On APImage it costs 42 credits per 5 seconds at 480p, 84 at 720p and 167 at 1080p, with reference clip seconds billed on top.

How much does Wan 3.0 cost?

42 credits per 5 seconds at 480p, 84 at 720p and 167 at 1080p, with sound included. Reference clips are billed on top of the output by their length. Prime speed costs more.

How long can a Wan 3.0 clip be?

From 2 to 30 seconds per generation.

What can I use as references?

Up to nine images, up to five video clips totalling 15 seconds at 16 fps or more, and up to five audio tracks totalling 15 seconds.

Can Wan 3.0 edit an existing video?

Yes. Add the clip as a reference and describe the change, or use it in the Recast tool to replace the lead while keeping the moves.

Which plan includes Wan 3.0?

Starter and up, out of the same credit pool as every other model.

How do I use it?

Open the video studio and pick Wan 3.0. Upload references: images, clips and audio. Write the shot and call each reference as @img1, @Video1 or @Audio1. Pick resolution, aspect ratio, length and speed, then generate.

Wan 3.0
is live.

Generate with Wan 3.0
From 42 credits per 5 secondsOne credit poolNo per-model surchargeEvery model, one plan