AlibabaAlibaba

Wan 3.0
30-Second AI Video With Native Audio

Wan 3.0 is Alibaba's next-generation video model, generating up to 30 seconds of video in a single continuous pass, with synchronized audio, across 480p, 720p, and 1080p. Direct scenes with text, images, video, audio, or even documents as reference.

Wan 3.0 video generation is available on faktry: text-to-video, image-to-video, and reference-to-video, all with native synchronized audio and clips up to 30 seconds.

Wan 3.0 in Action

See what Wan 3.0 can create, generated on faktry.

What Is Wan 3.0?

Wan 3.0 is Alibaba Tongyi Lab's next-generation video model, launched in public beta on August 6, 2026. It generates video from a text prompt, a still image, or reference material in a single continuous pass up to 30 seconds long, double the 15-second cap of Wan 2.7, with native synchronized audio generated alongside the video by default. Reference-to-video accepts up to 10 images, 5 video clips, and 5 audio clips, plus documents and public web pages as source material, folding reference, editing, replication, and motion-driving into one model.

Wan 3.0 at a glance

Developer
Alibaba (Tongyi Lab)
Released
Public beta, August 6, 2026
Type
Text-to-video, image-to-video, and reference-to-video model
Max video length
Up to 30 seconds per generation
Resolution
480p, 720p, 1080p
References
Up to 10 images, 5 videos, 5 audio clips, documents, web pages
Audio
Native, synchronized generation by default
On faktry
Available now

Last updated: · Model details published by Alibaba and fal.ai

Wan 3.0 vs Wan 2.7

What changes between Alibaba's established Wan 2.7 and the newer Wan 3.0.

CapabilityWan 2.7Wan 3.0
DurationUp to 15 secondsUp to 30 seconds
Resolution720p, 1080p480p, 720p, 1080p
AudioOptional driving audio uploadNative, synchronized audio by default
ReferencesReference images and videoImages, video, audio, documents, and web pages
EditingSeparate reference and edit endpointsReference, editing, replication, and motion-driving in one model
Aspect ratios16:9, 9:16, 1:1, 4:3, 3:4Adaptive auto-select, plus 16:9, 9:16, 1:1, 4:3, 3:4

Explore all AI models on faktry →

Wan 3.0, Live on faktry

Three ways to generate video with Wan 3.0, each with native audio and billed per second by resolution.

Text to Video

Generate a video with synchronized native audio directly from a text prompt.

3-12 credits / second, by resolution
2-30s · 480p, 720p, or 1080p · adaptive or fixed aspect ratio

Image to Video

Animate a starting image into a full video clip, with optional end-frame control for precise transitions.

3-12 credits / second, by resolution
2-30s · 480p, 720p, or 1080p · native audio

Reference to Video

Condition generation on up to 10 images, 5 videos, and 5 audio clips to keep characters and motion consistent.

3-12 credits / second, by resolution
2-30s · 480p, 720p, or 1080p · native audio

Duration, resolution, aspect ratio, and native audio are all configurable per generation. Lower resolutions cost less per second.

Why Choose Wan 3.0?

Alibaba's latest video generation technology, built for longer, richer, more controllable output.

Up to 30-Second Clips

Generate a full 30-second take in a single continuous pass, double the 15-second cap on Wan 2.7, for complex camera moves and unbroken shots.

480p, 720p & 1080p Output

Render fast, low-cost drafts at 480p or go straight to a clean 1080p final pass, priced per second at each tier.

Rich Multimodal References

Guide generation with up to 10 reference images, 5 video clips, and 5 audio clips, keeping characters, products, and motion consistent across a take.

Native Synchronized Audio

Audio, including multilingual voice, is generated alongside the video by default, instead of added as a separate pass.

Unified Reference & Editing

Reference, editing, replication, and motion-driving are folded into one model instead of separate specialized endpoints.

Sharper Motion & Faces

Improved motion smoothness, scene fidelity, and visual coherence, including realistic faces with synchronized micro-expressions.

Capabilities

A closer look at what Wan 3.0 can generate and control.

Video Generation

Text-to-video, image-to-video with optional end-frame control, and reference-to-video from images, video, audio, documents, and web pages
2 to 30 seconds per generation, in a single continuous pass
480p, 720p, or 1080p output, priced per second at each tier
Up to 10 reference images, 5 reference videos, and 5 reference audio clips per generation
Reference, editing, replication, and motion-driving unified in one model
Coherent camera movement and readable limb and contact mechanics across complex, full-body motion

Where Wan 3.0 Fits In

Longer, richer video generation for projects that used to need multiple takes or separate tools.

Long-Form Storytelling

Tell a complete scene in a single continuous 30-second take instead of stitching several short clips together.

Marketing Video

Produce longer, on-brand video content for ads and social campaigns with synchronized voice and sound baked in.

Product Visualization

Animate a product photo into motion, or pin a start and end frame to control exactly how a shot transitions.

Social Content

Generate reference-driven video for every platform, from a single source image, video, or document.

How It Works

Creating videos with Wan 3.0 is simple and powerful.

1

Describe Your Video

Write a detailed text description, or add reference images, video, audio, or a document to base the video on.

2

Configure Your Settings

Set duration, resolution, aspect ratio, and audio options. Add up to 10 reference images, 5 videos, and 5 audio clips for consistency.

3

Generate & Download

Get your video in minutes, with native synchronized audio already baked in.

Ready to Create with Wan 3.0?

Start generating next-gen videos today with free credits.

Free credits to try us

ACTIVE
100 credits included
Access to Wan 3.0 generation
All resolutions and aspect ratios
Commercial usage rights
Get Started Now

Frequently Asked Questions

What is Wan 3.0?

Wan 3.0 is Alibaba Tongyi Lab's next-generation video model, generating video with native synchronized audio from text, images, or reference material in a single continuous pass up to 30 seconds long.

What's the difference between Wan 3.0 and Wan 2.7?

Wan 3.0 doubles the maximum clip length to 30 seconds, adds a 480p tier alongside 720p and 1080p, generates native audio by default instead of requiring an uploaded audio track, and unifies reference, editing, replication, and motion-driving in one model.

How long can Wan 3.0 videos be?

Wan 3.0 generates clips from 2 to 30 seconds in a single continuous pass, double the 15-second cap of Wan 2.7.

What reference inputs does Wan 3.0 support?

Up to 10 reference images, 5 reference videos, and 5 reference audio clips, plus documents like PDF, Word, Excel, and PowerPoint files, and public web pages, to keep characters, products, and style consistent.

What resolutions does Wan 3.0 support?

480p, 720p, and 1080p, each priced per second, so you can render fast, low-cost drafts or a clean 1080p final pass.

Can I use Wan 3.0 videos commercially?

Yes! All videos generated with Wan 3.0 on faktry can be used for commercial purposes including marketing, advertising, and content creation.