Wan 3.0 Video

by AlibabaVideo

Alibaba Wan 3.0 generates native video up to 30 seconds from text, first and last frames, reference images, video, audio, webpages, or document files, with optional synchronized audio.

InputMultimodaltext · media · files
Duration2–30 secor smart duration
FlowAsynctask result URL
Price3.5 q/s480P promotional rate
Example clipVideo preview
Graduation close-up

Front-facing close-up of a young Black woman in a blue graduation gown. She speaks directly toward camera with a serious, worried expression, then gradually lowers her gaze toward what she is holding. Keep the background softly defocused, preserve natural skin texture under overhead light, and add only the subtle breathing movement of a handheld camera.

Text to video · 1080P · 10 seconds · With audio
Examples

Wan 3.0 creator examples

01 · Graduation close-upT2V
Text-to-video · no reference image
Prompt

Front-facing close-up of a young Black woman in a blue graduation gown. She speaks directly toward camera with a serious, worried expression, then gradually lowers her gaze toward what she is holding. Keep the background softly defocused, preserve natural skin texture under overhead light, and add only the subtle breathing movement of a handheld camera.

02 · Hospital corridorT2V
Text-to-video · no reference image
Prompt

A modern suspense-film long take with a cold pale-green grade. Beneath harsh fluorescent light, an empty hospital corridor has low saturation and extreme contrast. A woman in a dark green jacket bursts through a door at the far end and runs along the center line while the camera tracks backward. She reaches the nurses' station, slams both hands onto the counter, looks around desperately, and calls for help while her voice echoes through the nearly silent corridor.

03 · Boxing trainingT2V
Text-to-video · no reference image
Prompt

A 30-second Japanese sports-anime sequence inside a boxing gym. A short-haired athlete repeatedly strikes a heavy bag; each impact produces clear vibration and recoil. He is sweating, visibly exhausted, and fighting through pain as though trapped in a difficult moment.

04 · California shopping-cart one-takeT2V
Text-to-video · no reference image
Prompt

Create a 30-second, high-intensity cinematic one-take in a West Coast street-fantasy style, blending the rough texture of a 1990s skate tape with a polished modern commercial. A rebellious teenager rides inside a red metal shopping cart down a steep palm-lined California road. The camera tracks inches above the asphalt, passes beneath the airborne cart, drives straight into a giant billboard, and turns the rider into a flat torn-paper artwork before returning to street level to reveal the real teenager looking up at his frozen image.

05 · Legacy hardwareT2V
Text-to-video · no reference image
Prompt

Generate a 30-second adult science-fiction war animation with photoreal CGI, biomechanical soldiers built from pale bone and black industrial machinery, insect-like joints, abandoned future city streets, cold gray daylight, and restrained toxic-green life. The camera races through an automated battle before one skeletal soldier notices a tiny translucent green flower in cracked concrete, disobeys the red ADVANCE command, and shields the flower from incoming fire. Keep the tone oppressive and physical, with realistic destruction and no heroic fantasy gloss.

06 · Deep-sea emergenceT2V
Text-to-video · no reference image
Prompt

Create a 30-second photoreal disaster-film sequence at night. A small weathered fishing boat marked WAN struggles through violent rain and huge waves while an unnatural shadow approaches beneath the surface. Build tension through deck-level close-ups, frightened crew reactions, abnormal swells, a half-submerged camera, and brief lightning flashes. At the climax, an enormous wet, heavily textured deep-sea creature erupts beside the boat, throwing a wall of water into the storm with a convincing sense of scale and weight.

Capabilities

What you can build with Wan 3.0 Video

Native 30-second generation

Generate integer durations from 2 to 30 seconds, or let Wan select a smart duration for the prompt.

Multimodal references

Guide a result with images, video, audio, a webpage, or a document while preserving reference details and timing.

First and last frame control

Constrain the opening frame, or provide both opening and closing frames for a controlled transition.

Synchronized audio

Create video with an audio track by default, or disable audio for silent production workflows.

Pricing

Per-action, in quota

Usage is metered in quota by model, action, and output settings.

Wan 3.0 official channel · limited-time 30% offOfficial pricing
ModelResolutionTTAPI priceOfficial price
wan3.0-video480P3.5 quota / second5 quota / second
wan3.0-video720P7 quota / second10 quota / second
wan3.0-video1080P14 quota / second20 quota / second
API request

Wan 3.0 Video operations and request fields

Use the endpoint below with your TTAPI key and the request fields shown.

Text to videoPOST
/wan/api/v1/services/aigc/video-generation/video-synthesis

Generate a Wan 3.0 video from a text prompt without media input.

Official reference
First framePOST
/wan/api/v1/services/aigc/video-generation/video-synthesis

Generate a video while strictly preserving a supplied first frame.

Official reference
First and last framesPOST
/wan/api/v1/services/aigc/video-generation/video-synthesis

Generate a transition constrained by supplied first and last frames.

Official reference
Reference mediaPOST
/wan/api/v1/services/aigc/video-generation/video-synthesis

Generate from reference images, video, audio, or a web link.

Official reference
Reference filePOST
/wan/api/v1/services/aigc/video-generation/video-synthesis

Generate from a referenced document file that Wan interprets as creative context.

Official reference
FetchGET
/wan/api/v1/tasks/{task_id}

Retrieve task status, generated video URL, and usage by task ID.

Official reference
POSThttps://api.ttapi.io/wan/api/v1/services/aigc/video-generation/video-synthesis

Headers

TT-API-KEYstringrequired

Your TTAPI API key.

Content-Typestringrequired

Use application/json.

Body

modelenumrequired

Use wan3.0-video for the standard Wan 3.0 model.

input.promptstringconditional

Text prompt describing the video. Supports Chinese or English up to 20,000 characters; either a prompt or media input is required.

input.mediaobject[]conditional

Optional media inputs for first/last frames, reference images, reference video, reference audio, a web link, or a document file.

input.media[].typeenumconditional

first_frame, last_frame, reference_image, reference_video, reference_audio, link, or file. Reference media and file/link modes cannot be mixed in one request.

input.media[].urlstringconditional

Public URL or supported Base64 data for the corresponding media item.

parameters.resolutionenumoptional

Output resolution: 480P, 720P, or 1080P. Defaults to 1080P.

parameters.ratioenumoptional

adaptive, 16:9, 4:3, 1:1, 3:4, or 9:16. Defaults to adaptive.

parameters.durationintegeroptional

Output duration from 2 to 30 seconds, or -1 for smart duration. Defaults to 5 seconds.

parameters.audiobooleanoptional

Generate synchronized audio with the video. Defaults to true.

parameters.seedintegeroptional

Seed from 0 to 2,147,483,647 for reproducible generations.

parameters.prompt_extendbooleanoptional

Enable automatic prompt enhancement. Defaults to true.

parameters.watermarkbooleanoptional

Add a watermark when true. Defaults to false.

Open the official TTAPI documentation
Integrate

From key to first result

Copy the resolved endpoint, authentication headers, and request body for this model.

curl --request POST \
  --url https://api.ttapi.io/wan/api/v1/services/aigc/video-generation/video-synthesis \
  --header 'TT-API-KEY: $TTAPI_KEY' \
  --header 'Content-Type: application/json' \
  --data '{"model":"wan3.0-video","input":{"prompt":"Front-facing close-up of a young Black woman in a blue graduation gown. She speaks directly toward camera with a serious, worried expression, then gradually lowers her gaze toward what she is holding. Keep the background softly defocused, preserve natural skin texture under overhead light, and add only the subtle breathing movement of a handheld camera."},"parameters":{"resolution":"1080P","ratio":"16:9","duration":10,"audio":true,"seed":12345,"prompt_extend":true,"watermark":false}}'