Wan 2.5 is here

Wan 2.5
Fast, Efficient Video with Native Audio

Create native-audio AI video efficiently with the Wan 2.5 Video Generator.

Start Directing
0/1000
Create

What Is Wan 2.5

Wan 2.5 is Alibaba’s AI video model, and one of the model options built into DirectStudio’s video generator (powered by Wavel AI). It’s built for native synchronized audio, strong prompt adherence, and efficient rendering, so you get high-quality output without a long wait or a high credit cost.

Wan 2.5 pairs dialogue, ambient sound, and video generation in a single pass, while staying fast and cost-efficient enough for high-volume iteration — making it a strong fit for testing multiple hooks or variants before committing to a final render.

Key Features of Wan 2.5

Native audio, generated in sync

Synchronized dialogue and soundSpeech, ambience, and effects generated alongside the video in one pass.
Accurate lip-syncDialogue timing matches mouth movement without a separate sync step.
Audio + video.Generated together in a single pass, not layered on after.

Examples

Wan 2.5 example output
Veo 3.1 example output
Kling 3.0 example output
Seedance 2.0 example output
Sora 2 example output
Hailuo example output
Start creating

Trusted by teams, just like yours.

msnHubSpotNoBrokerByteDanceWondriumteachooEmeritusSpectrumIllinoisVerizon

Working with Wan 2.5

Limitations, and how
to get around them

Model choice takes judgment

Wan 2.5 isn't the automatic best pick for every video. DirectStudio offers several models because content type matters. What helps: match the model to the job; Wan 2.5 for fast iteration, native audio, and cost-efficient batches.

Draft Your Script
Chat prompt asking DirectStudio AI to storyboard a dialogue scene, with the AI confirming the drafted shots

Photoreal detail trails the heaviest flagship models

For maximum cinematic detail on a single hero shot, a heavier model can edge it out. What helps: use Wan 2.5 for volume and iteration, then re-render your final pick on a flagship model if needed.

Build a Storyboard
Editing a scene prompt with attached reference image and voice sample before rendering

Native duration is still capped per generation

A single pass tops out before needing to chain additional clips for longer content. What helps: plan longer videos as multiple generations extended together rather than one continuous render.

Open the Studio
Videos tab showing the full video and individual scene cards with Play all and Export Final Video actions

Use cases, with more DirectStudio tools

How To Use Wan 2.5

1

Pick your video type

Choose from faceless, AI avatar, UGC-style ad, commercial, cartoon, or music video.

2

Choose Wan 2.5 as your model

Alongside Seedance 2.5, Veo 3.1, Kling 3, and Seedream 5.

3

Upload media and enter your script

Add reference images, write your prompt, include dialogue where the scene calls for it.

4

Generate and review

Preview with synced audio, fine-tune in the editor, and export.

Start creating
DirectStudio storyboard workspace with Wan 2.5 selected

Where Wan 2.5 Excels

Professional video for every use case

High-volume ad testing exampleHigh-volume ad testingFast, cost-efficient generation makes it practical to test many hooks and variants.
Dialogue-driven scenes exampleDialogue-driven scenesNative audio and lip-sync for talking-head and presenter content.
Multi-part instruction-heavy prompts exampleMulti-part instruction-heavy promptsStrong prompt adherence keeps complex shot directions accurate.
Rapid prototyping exampleRapid prototypingIterate a storyboard quickly before committing credits to a final flagship render.
Start creating

Discover more

Text to video Script to video Image to video AI UGC Ads AI Ad generator Seedance 2.5 Kling 3 Veo 3.1 Seedream 5

A video studio built like a chat.

DirectStudio chat-driven video workspace
Don't stare at a blank canvas.Just tell the DirectStudio agent what you need, upload your brand assets, and let the AI structure your narrative.
Watch your story come to life in real-time.DirectStudio generates high-fidelity keyframes and an editable script artifact in seconds.
Drag and drop scenes.Rewrite prompts to regenerate specific clips, then export your final masterpiece.
Try DirectStudio for free

Pricing

Self-serve credits for every story workflow.

Use DirectStudio with pay as you go credits, fixed plans, and top-ups whenever you need more generation capacity.

Pay as you go

Top-ups
$0.20 per credit
Buy only what you need

Start self-serve and add credits only when your story pipeline needs more generation time.

  • Top AI ModelsSeedDance 2, Seedance 2.5, Veo 3.1, Kling 3, Nano Banana Pro, SeedDream 5, WAN 2.5
  • No monthly commitment
  • Use credits across AI models
  • Top up any time
Get started

Pro

$40/Month
$0.15 per credit
300 credits included

A focused starter pack for testing story ideas, producing first drafts, and iterating with prompts.

  • Top AI ModelsSeedDance 2, Seedance 2.5, Veo 3.1, Kling 3, Nano Banana Pro, SeedDream 5, WAN 2.5
  • Video Agent access
  • Prompt-based edits
  • Consistent characters and scenes
Get started
Best value

Scale

$100/Month
$0.12 per credit
1,000 credits included

Best value for regular creators producing more scenes, versions, and final video exports.

  • Top AI ModelsSeedDance 2, Seedance 2.5, Veo 3.1, Kling 3, Nano Banana Pro, SeedDream 5, WAN 2.5
  • Best credit rate
  • More model runs
  • Built for ongoing story production
Get started
Self serveCreate an account, add credits, and start producing without talking to sales.
Shared walletCredits apply across video, image, audio, edits, and final render workflows.
Top-upsAdd more credits whenever you need extra capacity for a bigger story batch.

Key features included

Video Agent
Top video, image and audio models
Edit with prompts
Characters, location, and costumes consistency
Pay as you go
Top-ups available anytime

Have any question?

Alibaba's AI video model, built for native synchronized audio, strong prompt adherence, and efficient, cost-effective rendering.

Yes. Dialogue, ambient sound, and effects are generated in the same pass as the video, no separate audio step needed.

It trades a small amount of maximum photoreal detail for faster, more cost-efficient generation, making it well suited to high-volume iteration and testing.

Yes, alongside text-to-video generation — feed a reference image and prompt to animate it.

It's available as one of the selectable video models inside DirectStudio's video generator, alongside Seedance 2.5, Veo 3.1, Kling 3, and Seedream 5.

DirectStudio runs on credits. Pay as you go, or subscribe to Pro or Scale for a lower per-credit rate and more included generation capacity.

Still have a question in mind?

Contact us if you have any other questions.

Contact us

Wan 2.5 is live

Native audio, strong prompt adherence, and efficient rendering — available now.

Start Directing Free