VidCella LogoVidCella

Wan 3.0: 30-Second AI Video with Native Audio

Alibaba's all-in-one video model, now on VidCella. Up to 30 seconds in a single shot, 1080p, native audio, and first/last frame control. From 15 credits/s.

Loading...
Key Capabilities

What Wan 3.0 Does Differently

One model, three ways in — and a 30-second ceiling most models can't reach.

First & Last Frame Control

Give it a starting frame and Wan 3.0 animates from there. Add a last frame and it fills the motion between the two, so you decide where the shot begins and ends.

Multimodal Reference

Up to 10 images, 5 videos and 5 audio clips in one generation. Name them in the prompt as Image 1, Video 1, Audio 1 and assign each a role in the scene.

What Is Wan 3.0?

Alibaba's latest video model — text, images, video and audio in, up to 30 seconds of 1080p video with synchronized native audio out.

  • 30 Seconds, One Shot
    No stitching. A single generation runs 2 to 30 seconds at 480p, 720p or 1080p.
  • Native Audio
    Score, dialogue and ambience generated with the picture, timed to the cut.
  • Three Ways In
    Text to Video, Image to Video with first/last frame control, or Reference to Video built from your own images, clips and audio.
  • Normal or Prime
    Prime is the faster channel at a higher rate. Same options, same output quality.

How to Use Wan 3.0 on VidCella

Four steps from prompt to finished clip:

Wan 3.0 Features on VidCella

Full capability surface, pay-per-generation:

Up to 30 Seconds

2-30s in a single generation, at 480p, 720p or 1080p.

Native Audio

Soundtrack, dialogue and ambience generated in the same pass.

Three Input Modes

Text, image with first/last frame, or full multimodal reference.

Multimodal Reference

10 images + 5 videos + 5 audio clips per generation.

First & Last Frame Control

Set where the shot begins and ends; Wan 3.0 fills the motion between.

Normal & Prime Speed

From 15 credits/s. Prime generates faster at a higher rate.

FAQ

FAQs about Wan 3.0

Common questions about Wan 3.0 on VidCella

1

What is Wan 3.0?

Alibaba's latest video model, in public beta since August 2026. It takes text, images, video and audio and returns up to 30 seconds of video with synchronized native audio.

2

How does Reference to Video work?

Upload up to 10 images, 5 videos and 5 audio clips, then cite them in the prompt as Image 1, Video 1, Audio 1 and give each a role in the scene. Reference videos count toward the 30-second limit and the billed duration.

3

What's the difference between Normal and Prime?

Prime is a faster upstream channel. The options and output quality are the same — you're paying for turnaround, not for better video. Prime costs 25% more on Text and Image modes and 50% more on Reference.

4

How long can a Wan 3.0 video be?

2 to 30 seconds in a single generation, with no stitching. Reference videos count toward the total, so their length plus your output must stay within 30 seconds.

5

What resolutions and aspect ratios are supported?

480p, 720p and 1080p, in 16:9, 9:16, 1:1, 4:3 or 3:4. Image to Video follows your first frame's ratio unless you pick one.

6

How much does Wan 3.0 cost?

Per second, no subscription. Text and Image to Video: 15 / 30 / 60 credits per second at 480p / 720p / 1080p. Reference to Video is the same on Normal, and reference videos add their length to the billed duration.