Omni-reference video model

WAN 3.0.

WAN 3.0 from Alibaba brings text, image, reference and native-audio video generation into one model. Build cinematic tests, product reveals and social cuts up to 30 seconds - with the credit meter visible before you render.

By Alibaba 480p / 720p / 1080p Up to 30 seconds Native audio Omni-reference inputs
Live route pricing

Per-second pricing that stays visible.

Choose the resolution first, then see the route rate before you render. The public values below refresh weekly.

1080p

Highest detail
32 credits per second ~ $0.16 per second
  • Best fit for hero shots and final reviews
  • Native audio route where enabled
  • Cost scales directly with clip length
Create with WAN 3.0

720p

Balanced
16 credits per second ~ $0.09 per second
  • Useful for fast concept routes
  • Same prompt and reference workflow
  • Lower-cost option testing
Try the balanced route

480p

Fastest iteration
8 credits per second ~ $0.04 per second
  • Lowest-cost storyboard passes
  • Good for early motion exploration
  • Upgrade the resolution after approval
Test a cheaper route

Costs are shown again inside the studio before each render. Resolution, length and rerenders all show their credit impact upfront.

What WAN 3.0 can do

Six things WAN 3.0 does for working creators.

Omni-reference

Keep the right visual references in frame

Use reference images and video context to guide characters, products and environments while the prompt directs the action.

First and last frame

Control where the motion starts and ends

Give the shot a visual boundary instead of hoping the model lands on the right final beat - useful for transitions, reveals and match cuts.

Native audio

Dialogue, ambience and effects in one render

Turn on native audio when the shot needs synchronized speech, atmosphere or effects alongside the picture.

Longer clips

Take a scene through a full 30-second beat

WAN 3.0 supports clips up to 30 seconds, so a single generation can cover more than a micro-moment when the brief needs room to breathe.

Text-to-video

Go from a written brief to a moving scene

Describe the camera, subject, motion and sound in one prompt. WAN 3.0 handles the first visual pass without a separate storyboard render.

Image-to-video

Bring a still frame to life

Start from a product shot, character design or environment board and ask WAN 3.0 to build the motion around it.

Spec sheet

The numbers, plainly.

Resolution

480p, 720p and 1080p routes. Choose the resolution that fits the brief and the budget.

Aspect ratios

Wide, vertical and square framing are supported for cinematic, social and product work.

Clip length

5 to 30 seconds with per-second pricing shown before you render.

Audio

Optional native synced audio for dialogue, ambience and effects in the same render.

Inputs

Text prompts, first-frame image-to-video, first/last-frame control and omni-reference workflows.

Pipeline

Built by Alibaba. You see one FCKexpensive.AI credit cost before dispatch.

Where it earns its credits

WAN 3.0 is the right tool when…

Reference-led product films

Keep a product, character or environment visually anchored while you explore movement and camera direction.

Pitch & concept films

Use longer clips and native audio when a client pitch needs to feel closer to a finished scene.

Product reveals from stills

Turn packshots, character boards and environment art into moving shots for campaign exploration.

Vertical social-first edits

Start with a 9:16 brief and native audio for social-first concepts ready to review and iterate.

Try WAN 3.0 with your 150 free sign-up credits.

The credit meter stays visible: choose your duration, check the resolution rate, then decide whether to render.