Wan 3.0 AI Video Generator
Wan 3.0 is Alibaba Cloud Model Studio's all-in-one Wan video generation family, with up to 30-second output, native audio, 480P–1080P resolution, first/last-frame control and multimodal reference workflows.
WanVideoMaker is an independent Wan information resource. Current generation is available through Ikat, our multi-model AI video studio.
What is Wan 3.0?
Wan 3.0 consolidates multiple video-generation workflows into the same current model family. Official guidance lists text-to-video, first-frame image-to-video, first-and-last-frame generation, multimodal reference generation, video editing and extension.
The important change is not only longer output. Wan 3.0 can combine visual, audio and reference information in one workflow, which expands the model's usefulness for character consistency, product consistency, dialogue-driven scenes and longer narrative shots.
Supported Wan 3.0 workflows
These workflow categories are documented by Alibaba Cloud. Exact input limits can change, so production integrations should always verify the current official guide.
Text to video
Generate video from a text prompt with controllable resolution, aspect ratio, duration and audio settings.
First-frame image to video
Use an image as the opening frame and animate it into a generated video.
First + last frame
Provide opening and ending frames to guide the transition and final composition.
Multimodal reference
Use images, video clips and audio as references for identity, motion, scene, voice or style.
Document and webpage reference
Use supported files or a public webpage as reference material in supported generation workflows.
Current official guidance supports up to 20 multimodal reference materials in supported workflows, including up to 10 images, 5 video clips and 5 audio clips, subject to per-input limits.
What changed with Wan 3.0?
Up to 30 seconds
Longer single generations create more room for complete actions, camera movement and narrative pacing.
Native audio-visual generation
Official materials describe dialogue, background music and sound effects generated together with the video.
Multimodal reference
Images, video, audio and supported documents or webpages can participate in supported workflows.
One model family
Standard and Prime cover the current all-in-one workflow family, with Prime positioned for faster inference.
Wan 3.0 Standard vs Prime
| Model | Best fit | Pricing scope |
|---|---|---|
wan3.0-video | General Wan 3.0 generation | Per-second rate by output resolution |
wan3.0-video-prime | Faster inference when turnaround matters | Separate regional pricing |
Prime is intentionally excluded from our Standard calculator because Alibaba Cloud publishes separate regional Prime prices.
Wan 3.0 pricing at a glance
The calculator uses Alibaba Cloud Model Studio International/Singapore Standard rates in USD.
Alibaba Cloud's current pricing documentation states that billable duration can include input video duration plus successfully generated output video duration. Without input video, billable seconds equal the generated output duration.
The current international 30% launch offer is published through September 24, 2026, 00:00 (UTC+8); final charges remain subject to the provider console and bill.
Open the full Wan 3.0 pricing calculatorHow to use Wan 3.0
- 1
Choose the workflow
Start with text, a first frame, first + last frames, or supported reference material.
- 2
Set output controls
Choose resolution, ratio and duration. Supported workflows can also use adaptive ratio or smart duration.
- 3
Generate and iterate
Refine prompts and references, then compare outputs instead of assuming one prompt will always produce the final result.
Wan 3.0 prompt basics
A useful Wan prompt usually separates subject, action, environment, camera behavior, lighting or visual style, and audio intent. When reference media is used, describe what each reference should contribute instead of repeating every visual detail in prose.
A dedicated prompt library should be expanded from real Search Console demand rather than mass-publishing thin prompt pages.
Wan 3.0 FAQ
What is Wan 3.0?
Wan 3.0 is Alibaba Cloud Model Studio's current all-in-one Wan video generation family. The Standard model ID is wan3.0-video, and wan3.0-video-prime is a faster-inference variant.
How long can Wan 3.0 videos be?
Current Alibaba Cloud documentation lists generated output durations from 2 to 30 seconds for Wan 3.0.
What resolutions does Wan 3.0 support?
Current official documentation lists 480P, 720P and 1080P output at 30fps.
Does Wan 3.0 generate audio?
Yes. Alibaba Cloud describes native audio-visual generation that can produce dialogue, background music and sound effects, with controls depending on the workflow.
Can Wan 3.0 use images, videos or documents as references?
Yes. Current guidance supports multimodal reference workflows using images, video clips, audio, supported files and public webpage links, subject to input limits.
How is Wan 3.0 billed?
Alibaba Cloud bills by video duration and states that billable duration can include input video duration plus successfully generated output video duration. Workflows without input video naturally bill only output video seconds.
Is Wan 3.0 free?
Wan 3.0 API usage is billed under current Alibaba Cloud Model Studio pricing. Some provider regions publish limited free quota, but WanVideoMaker does not claim permanent free generation.
Official Wan 3.0 sources
Capability and pricing claims on this page are anchored to Alibaba Cloud Model Studio documentation and are date-stamped so time-sensitive details can be rechecked.
Ready to create with Wan 3.0?
Use this page for verified model context, then continue to Ikat when you want to generate with Wan in a current multi-model workflow.