Can Wan 2.5 turn both text and images into video?

Yes. Use Text to Video to build a scene from a written prompt, or Image to Video to animate a first-frame image while keeping its composition and visual identity as the starting point.

How long can a Wan 2.5 video be?

Wan 2.5 supports clips up to 10 seconds. The duration options currently available for your selected mode are shown directly in the generator.

Which resolutions and aspect ratios are available?

Wan 2.5 supports output up to 1080p. The provider workflow exposes settings by mode, including landscape, portrait, and square formats for text-to-video.

Does Wan 2.5 generate audio?

Yes. Alibaba describes Wan 2.5 as supporting native audio-visual synchronization, so motion and sound can be generated as an aligned short clip. Availability can depend on the selected provider route.

How many credits does Wan 2.5 use?

Credit use depends on mode, resolution, duration, and your plan. The generator calculates and displays the current credit requirement before you create.

Meta Muse Video

Alibaba Wan video generation

Wan 2.5 AI Video Generator

Turn a written scene or a single image into a polished video with coherent motion, strong prompt following, and native audio-visual synchronization.

1080pMaximum resolution

10 secMaximum clip length

Native A/VSynchronized sound

Generation Route

Prompt *

0 chars

Aspect Ratio

Duration

Resolution

Available Credits

See sample outputs for Text to Video in the Meta Muse Video workspace

Motion directions

From a Single Idea to a Moving Scene

Explore the kinds of product motion, character moments, and imaginative scenes you can build with a focused video prompt or first-frame image.

00:05

Product Motion

Turn a clean product frame into an atmospheric commercial shot with deliberate camera movement.

00:10

Cinematic Portrait

Add natural expression, environmental motion, and camera language to a character-led scene.

00:05

Imaginative Worlds

Translate detailed visual direction into dynamic fantasy, VFX, and story-driven moments.

Creative workflow examples shown for inspiration. Generated results vary by prompt, input, and selected settings.

Wan 2.5 capabilities

Built for Coherent Motion and Sound

Wan 2.5 combines text, visual, and audio understanding to keep the action, camera, and soundtrack aligned across a complete short clip.

Scene directionSynchronized clip

0s2.5s5s7.5s10s

Prompt

Stronger Prompt Following

Describe subject, action, environment, camera, lighting, and mood in one production-ready brief.

10 sec

Up to 10 Seconds

Create a complete short-form beat with more room for action development and camera movement.

A/V

Native Audio Sync

Generate motion and sound as an aligned result instead of treating audio as a separate afterthought.

1080p

Up to 1080p

Choose the available resolution and aspect ratio that fit ads, social feeds, and cinematic layouts.

November 2025 release context

Three Upgrades That Defined Wan 2.5

Wan 2.5 was more than a routine quality refresh. Compared with the earlier Wan 2.2 video line, its launch connected sound to the generated scene, doubled the available shot length, and pushed control toward commercial production needs.

Read the Alibaba Cloud launch overview

01Sound becomes part of the scene

From Silent Output to Synchronized Audio-Visual Generation

Wan 2.2's standard text-to-video and image-to-video models produced silent clips. Wan 2.5 introduced synchronized audio generation, allowing motion and sound to be conceived as one result instead of two disconnected production passes.

Wan 2.2: Silent video followed by a separate sound workflow
Wan 2.5 launch: Picture and matching audio generated as an aligned clip

The release positioned audio-visual synchronization as a core architecture upgrade, not a post-production effect.

02A longer creative beat

Ten Seconds Gave Motion and Story More Room

Earlier Wan 2.2 general video routes were fixed at five seconds. Wan 2.5 added a ten-second option, making it easier to stage a complete camera move, product reveal, or short narrative action without immediately stitching clips together.

Wan 2.2: Five-second output for standard T2V and I2V
Wan 2.5 launch: Five or ten seconds at 30 fps

The longer window was available for both text-to-video and first-frame image-to-video workflows.

03Commercial-grade direction

Better Adherence, Motion, and Image Quality in the Same Upgrade

Alibaba's launch materials grouped stronger instruction adherence, improved motion, superior image quality, and more controllable multimodal input as one production-focused step forward—aimed at film, advertising, and repeatable commercial workflows.

Earlier workflow: More iteration to hold action, framing, and finish together
Wan 2.5 launch: Stronger direction following with more polished motion and visuals

The differentiator was the combination: control, motion, and finish improved together rather than as isolated settings.

These comparisons describe the model capabilities announced at launch. Controls exposed in this generator can vary with the connected provider route.

Where Wan 2.5 Fits Best

Use it when a still concept needs convincing motion, clear direction, and a complete audio-visual beat.

Product and Campaign Shots

Prototype launch films, pack shots, fashion moments, and branded social assets without a full production setup.

Short-Form Storytelling

Create vertical, square, or landscape story beats for social campaigns, teasers, and narrative sequences.

Image-to-Video Animation

Start with your approved keyframe, then add controlled subject motion, camera movement, and atmosphere.

A Clear Wan 2.5 Workflow

Give the model one readable scene, choose the delivery format, then refine the motion instead of rewriting everything.

Step 1

Start from Text or an Image

Write a complete scene prompt, or upload a strong first frame that establishes subject, composition, and visual style.

Step 2

Set Format and Timing

Choose the aspect ratio, resolution, and clip duration that match the destination and the amount of action.

Step 3

Generate and Refine

Review camera motion, subject consistency, and audio alignment. Tighten one instruction at a time for the next version.

Wan 2.5 for Everyday Video Production

Move from a static idea to a shareable short clip in one focused generation workflow.

Prototype Faster

Test visual concepts and camera directions before committing to a full shoot or long edit.

One prompt to motion

Start the Way You Think

Use a written scene when exploring, or a first-frame image when composition must stay anchored.

Text or image input

Keep Costs Visible

See the required credits for the selected resolution and duration before starting a video.

Credits shown before generation

Create Delivery-Ready Detail

Generate at the resolution that matches quick drafts, social publishing, or higher-fidelity review.

Up to 1080p

Fit Every Content Format

Plan compositions for widescreen stories, vertical feeds, and square campaign placements.

Landscape, portrait, or square

Keep the Beat Together

Use synchronized audio-visual generation to establish a more complete short-form scene.

Unified motion and sound

Wan 2.5 FAQs

Key model facts based on Alibaba Cloud documentation and the available provider workflow.

Wan 2.5 is Alibaba's video generation model family for text-to-video and image-to-video creation. It improves instruction following, motion quality, visual detail, and synchronized audio generation.

Turn Your Next Frame into Motion

Write the scene or upload a keyframe, choose the format, and create a synchronized short video with Wan 2.5.