
Start from a prompt
Wan 3.0 Text to Video
Turn a prompt into a video of up to 30 seconds at 30fps, with native dialogue, music and sound effects.
ALIBABA VIDEO MODEL FAMILY
Alibaba’s video family makes up to 30 seconds at 30fps with native dialogue, music and sound effects. Start from text, a still or references, then edit or extend the result, all through one SandBase key.
Videos of up to 30 seconds in a single run.
Smooth motion for social, ads and story clips.
Dialogue, music and sound effects generated with the picture.
Six workflows, each in a Standard and a Prime tier.
AVAILABLE WORKFLOWS
Each route keeps its own request schema, parameters and live price. Open a workflow to run it, or switch to Prime for the higher-quality tier.

Start from a prompt
Turn a prompt into a video of up to 30 seconds at 30fps, with native dialogue, music and sound effects.

Start from a still
Animate a photo or keyframe into a cinematic clip with natural motion, camera movement and optional audio.

Keep subjects consistent
Generate from reference images, videos and audio, keeping subjects, motion and voice consistent with the references.

One route for every input
Text, a first and last frame, or reference images, videos and audio in a single route, with native audio.

Change an existing clip
Restyle or modify an input video from a text instruction, optionally guided by reference images and audio, keeping its length.

Continue a clip
Continue an input video forward from a text instruction, optionally guided by reference images and audio, with native audio.
PROMPT GUIDE
Wan 3.0 generates audio with the video, so a strong prompt covers both. Change one control at a time, check the clip, then keep only what improved it.
State the aspect ratio, length and where the clip will run.
Name the subject, its key details, the location and the light.
Break a longer clip into short time ranges so actions happen in order.
Give shot size, movement and pace for each beat.
Describe dialogue (with the exact line), ambience and music separately from the picture.
End with what must not change: identity, product shape, wardrobe or setting.
COMMON QUESTIONS
Wan 3.0 is Alibaba’s video generation family. It creates videos of up to 30 seconds at 30fps with native dialogue, music and sound effects, and it can also edit or extend an existing video.
Prime is the higher-quality tier of the same workflows and costs more per run. Start with Standard to settle the prompt, framing and references, then render the final version on Prime if the extra quality matters for the delivery.
Use Text to Video for a prompt-led concept, Image to Video when a still should set the composition, and Reference to Video when a character, product or voice must stay consistent. Use Video Edit or Video Extend when you already have a clip.
Yes. Wan 3.0 routes generate native audio, including dialogue, music and sound effects, together with the picture. Describe the voice, ambience or music you want in the prompt, or supply reference audio where the route accepts it.
Each route lists its live price on its model page. Prices depend on the route and tier, so check the page for the workflow you plan to use before running a batch.
CONTINUE ON SANDBASE
See Wan 3.0 next to Seedance 2.5, Gemini Omni Flash and the rest of the video catalog.