№ 15 VIDEO MODEL · ALIBABA

Wan 3.0

VIDEO · 6S/15S · 480P–1080P · MULTIMODAL REFERENCES TEXT → VIDEO IMAGE → VIDEO

Wan 3.0 is Alibaba Cloud's all-in-one video model: clips from 2 to 30 seconds, three generation entries, and a reference system that spans images, video, audio, documents, and public webpages. The most versatile input surface in the Wan line.

What it can do

Model specs from Alibaba; prices are in Motheia credits.

STARTS FROM
Text to video · First/last-frame generation · file or webpage generation
QUALITY
480p · 720p · 1080p
CLIP LENGTH
6 or 15 seconds
PRICE
480p · 6s: 300 credits · 720p · 6s: 550 credits · 1080p · 6s: 1050 credits · 1080p · 15s: 2600 credits
HOW IT IS PRICED
Per generated clip, priced by quality and length.
NOTES
Reference images, video, and audio accepted; documents up to 100MB/50 pages and public webpages as inputs; output audio on by default

The studio shows the exact credit cost before you generate.

Core features

01

Thirty-second range

2–30 whole seconds per generation, the widest span in the Wan line.

02

Three generation entries

General prompts with references, first/last-frame control, or file and webpage inputs.

03

Document and webpage inputs

Word, PDF, slides, and public pages can ground a generation.

04

Default output audio

Audio is on by default from the developer and does not change its price.

Where it earns its credits

  1. 01

    Explainer drafts

    Turn a document or public page into a moving summary.

  2. 02

    Controlled transitions

    First-and-last-frame entry for precise opens and closes.

  3. 03

    Longer cuts

    Up to 30 seconds for complete short-form pieces.

How to use it

Three steps, one balance, no new account per model.

  1. 1.

    Describe the result

    Open the studio, write the scene, motion, and camera direction in plain language.

  2. 2.

    Choose Wan 3.0

    Pick Wan 3.0 in the model selector and choose quality and length; the exact credit cost appears before you commit.

  3. 3.

    Generate and track

    Submit once — the job stays attached to your account, and the result waits in your history even if you leave the page.

Wan 3.0 FAQ

i. What is Wan 3.0?

Alibaba Cloud's all-in-one AI video model, identified as wan3.0-video: 2–30 second clips with general, first/last-frame, and file-or-webpage generation entries.

ii. How much does Wan 3.0 cost on Motheia?

Pricing starts at 300 credits for a 480p · 6s clip, with 720p at 550, 1080p at 1050, and a 15-second 1080p option at 2600. The studio shows the exact cost before you generate.

iii. Can I mix reference media with first/last frames?

No. The developer treats frame inputs as one family and reference media as another; requests cannot mix them.

iv. Does output audio cost extra?

No. At the developer, audio is enabled by default and the price is unchanged if disabled.

v. What documents can ground a generation?

The developer lists Word, Excel, PowerPoint, PDF, TXT, Keynote, Pages, Numbers, and Markdown up to 100MB and 50 pages, plus public webpages. The studio accepts the inputs shown in its upload field.