Wan 3.0 Text to Video: Scripted Clips With Sound

Last verified: October 19, 2026

Wan 3.0 works best for short story-driven video with sound: an ad spot, a short-drama scene, a music-video moment, where one written brief has to carry the shots, the lines and the soundtrack. The only input is a prompt, and from it Wan 3.0 text-to-video can build a multi-shot story of up to 30 seconds with a synchronized soundtrack. Alibaba reports that the model is already used for short drama and film, advertising, tourism promotion and music videos.

Because you only give it text, your prompt is the shooting script. Alibaba's prompt formula for Wan 3.0 puts everything in the prompt: shots with start and end seconds, spoken lines, sound effects and music, style and mood. So your wording decides where the clip cuts and what the viewer hears.

Write that script in the Prompt box. If you only have a rough idea, Enhance Prompt can suggest fuller versions. Then set the clip's length with Video Duration, its sharpness with Resolution and its frame shape with Aspect Ratio. Generate Audio controls whether the soundtrack is made along with the picture.

Explore more text to video

Capability Snapshot

Wan 3.0 at a Glance

This generator's input, settings and output on one screen.

Input

A text prompt of up to 5,000 characters

Video Duration

3 to 30 seconds

Resolution

480P, 720P or 1080P

Aspect Ratio

Auto, 1:1, 3:4, 4:3, 9:16, 16:9

Audio

Switch on or off, on by default

Output

1 video per run

How Do You Write a Better Wan 3.0 Prompt?

Open With a Summary Line That States the Length

Alibaba's own text-to-video samples begin with a line that sets the running time, genre and tone. One opens with a 20-second ultra-widescreen plush-universe adventure and is sent with a duration of 20 seconds. Do the same: put length, genre and mood in your first sentence, and set Video Duration to the same number, so the pacing you describe matches the clip you get back.

Give Every Shot a Lens, Angle and Camera Move

In Alibaba's formula, each shot has four parts: subject, scene, motion and aesthetic control (how it's filmed). Its text-to-video example uses that last part for concrete camera choices, such as a miniature low-altitude POV skimming the ground and an ultra-wide angle that exaggerates scale. Replace vague words like cinematic with a shot size, a lens, an angle and one stated move per shot, so every shot has a camera decision written into it.

Close the Prompt With a List of Exclusions

Alibaba's full prompt formula ends with a negative prompt list, placed after style and mood. Make the last line of your Prompt a short list of things that must not appear, such as subtitles, logos, extra people or a costume change. Keep that line to exclusions only and don't repeat what you already asked for, so the main description stays clear.

How Do You Make a Video With Wan 3.0?

Four steps take you from a written scene to a clip you can download.

1

Write your scene in the "Prompt" box

The "Prompt" box is required and holds up to 5,000 characters. That's room for a summary line, numbered shots, dialogue and sound cues. If you're starting from a rough idea, click "Enhance Prompt" under the box to see fuller versions; your own text stays as it is unless you pick one. Finish the script before touching the settings, because the next step should match what it describes.

2

Set "Video Duration", "Resolution" and "Aspect Ratio"

"Video Duration" offers every whole second from 3 to 30 and starts at 3, so raise it to the length your script needs. "Resolution" offers 480P, 720P and 1080P, with 480P selected by default. Draft at 480P while you check timing and framing, then switch to 1080P for the version you'll publish. "Aspect Ratio" starts on 1:1 and also offers Auto, 3:4, 4:3, 9:16 and 16:9. Pick 9:16 for vertical feeds or 16:9 for widescreen before generating, because the frame shape affects how every shot is composed.

3

Click "Generate"

First check the "Generate Audio" switch. It's on by default, so the clip comes with a soundtrack unless you turn it off. Turn it off if you plan to score or voice the video yourself, and leave it on for any prompt with dialogue, since the spoken lines are part of the audio. Then click "Generate" to start the run.

4

Download the result

Each run returns one video. Play it with sound before you download: check that the cuts land where you wrote them and that no unrequested text has appeared. Download it once it matches your script. If it doesn't, adjust the prompt and run it again.

Fixing Common Wan 3.0 Problems

What usually goes wrong with a text-only Wan 3.0 run, and what to change before the next one.

I asked for one continuous shot, but the video cuts to another angle.

Wan 3.0 can turn a text prompt into one unbroken shot or into a sequence of cuts, so if the prompt never says which, the model decides where to cut. Alibaba's own one-take text-to-video example asks for one continuous shot right in its opening line. Put that request (for example, single continuous take, no cuts) at the top of your Prompt, remove any shot numbers or timestamps, and generate again.

Signs, titles or captions in the video come out as garbled letters.

Alibaba names on-screen text rendering accuracy as an area where Wan 3.0 still has room to improve, so lettering generated inside the clip can come out misspelled or scrambled. Leave readable text out of the scene, ask for no subtitles or on-screen text, and add titles in your editor afterwards. If stray letters still show up, run the prompt again.

My clip has music or talking I never asked for.

With audio on, Wan 3.0 generates dialogue, background music and sound effects for a text prompt by itself, so if the prompt says nothing about sound, the model chooses the mix. Spell out what you want, the way Alibaba's own text-to-video example asks for no dialogue, no BGM and only ambient sound. Or switch off Generate Audio if you'll add your own track. Alibaba also lists audio texture as an area it is still improving, which is another reason to replace the soundtrack for a final cut. Then run it again.

My full shot list won't fit in the prompt box.

A detailed script with dialogue and sound for every shot can go past the Prompt box's character limit. Alibaba's prompt formula is meant to be trimmed: write the parts you want to control and skip the rest. Describe the character and setting once in the opening line, keep each shot to framing, action and its sound, and cut adjectives that repeat from shot to shot. Paste the shorter script back in and generate.

The video ends after just a few seconds.

Video Duration starts at its shortest option. If you leave it there, the run returns a very short clip, however much action the prompt describes. Raise Video Duration to the length your script needs before you click Generate, then run the prompt again.

Frequently Asked Questions

Can Wan 3.0 make a video longer than 30 seconds?

Not in a single run. Alibaba lists 30 seconds as Wan 3.0's longest single generation, and Video Duration stops at that same length, so anything longer has to be built from several runs. Split the story into scenes that each fit in one clip, use the exact same character and setting descriptions in every prompt, and join the clips in your editor.

How do I make characters speak in a Wan 3.0 video?

Write the line into the prompt and say who speaks it. Alibaba's formula has a dialogue slot written as the speaker, the word says, and the line in quotation marks. Put the delivery right next to it, such as low, hurried or warm. Give sound effects their own sentences, with a time range if you want them at a specific moment. Keep Generate Audio on so the speech is included in the clip.

Why does one Wan 3.0 video cost more than another?

The cost of a run is calculated from the options you choose, and the two that change it are Video Duration and Resolution. Once you're signed in, the amount for your current settings is shown above the Generate button before you start, so you can see what a longer clip or a higher resolution does to it before you commit. Enhance Prompt suggestions don't use credits.

How do I write a multi-shot prompt for Wan 3.0?

Start with one sentence that sums up the whole video. Then number each shot, put its time range right after the number (Shot 1 [0–3 s], Shot 2 [3–6 s] and so on), and follow it with what happens in that shot. The shots should run end to end with no gaps or overlaps. Put each shot's camera, action and dialogue inside that shot rather than in a separate paragraph, so every instruction is tied to a time.

Can Wan 3.0 make animated or stylized videos?

Yes. One of Alibaba's own text-to-video samples is a 3D animated short film with a dark-fantasy theme, shot as a single take, and its prompt names that style first. Put your style in the prompt's opening line the same way, so it applies to the whole clip instead of a single shot.

References

Sources and citations used to support the content provided above.

Updated: 2026-10-19 23:41:58 6 Sources

docs.modelstudio.console.alibabacloud.com

Source Link
https://docs.modelstudio.console.alibabacloud.com/en/model-studio/wan3-video-generation-guide

money.usnews.com

Source Link
https://money.usnews.com/investing/news/articles/2026-08-24/alibaba-launches-wan3-0-ai-video-model-after-10-billion-share-sale

www.alibabacloud.com

Source Link
https://www.alibabacloud.com/help/en/model-studio/wan3-video-generation-prompt-guide

help.aliyun.com

Source Link
https://help.aliyun.com/en/model-studio/wan3-video-generation-prompt-guide

www.alibabacloud.com

Source Link
https://www.alibabacloud.com/help/en/model-studio/wan3-video-generation-guide

www.alibabacloud.com

Source Link
https://www.alibabacloud.com/blog/wan3-0-30-second-ai-video-generation-from-any-input_603452