Video

Vidu Q3

Long-form video with native audio generated in a single pass

About model

On Together AI, Vidu Q3 costs $0.0975 per video.

Vidu Q3 is ShengShu Technology's flagship video generation model, producing up to 16 seconds of 1080p video at 24 fps with native audio, including dialogue, sound effects, and music, generated in the same pass as the visuals. It supports text-to-video, image-to-video, start and end frame control, and reference-to-video from up to four reference images, with intelligent camera switching that keeps subjects consistent across camera positions. Available on Together AI.

‍

Max Clip Length

16s

Long-form clips in a single generation

Output Resolution

1080p

At 24 frames per second

Audio-Video Generation

Native

Dialogue, effects, and music produced with the visuals

Model key capabilities

Native Audio-Video: Dialogue, sound effects, and music generated in the same pass as the visuals
Long-Form Generation: Up to 16 seconds of 1080p video in a single generation
Frame & Reference Control: Start and end frame control, plus reference-to-video from up to four images
Available on Together AI: Serverless access through the Together AI API

API usage

cURL
Python
Typescript

Endpoint:

vidu/vidu-q3

curl --request POST \
  --url https://api.together.ai/v2/videos \
  --header "Authorization: Bearer $TOGETHER_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
    "model": "vidu/vidu-q3",
    "prompt": "some penguins building a snowman"
  }'

from together import Together

client = Together()

# Create a video generation job
job = client.videos.create(
    prompt="A serene sunset over the ocean with gentle waves",
    model="vidu/vidu-q3"
)

import Together from "together-ai";

const together = new Together();

async function main() {
  // Create a video generation job
  const job = await together.videos.create({
    prompt: "A serene sunset over the ocean with gentle waves",
    model: "vidu/vidu-q3"
  });

Model card
Architecture Overview:
• ShengShu Technology video generation model in the Vidu Q3 family
• Text-to-video, image-to-video, start and end frame, and reference-to-video from up to four reference images
• Up to 16 seconds of 1080p video at 24 fps with native audio generated in the same pass

Training Methodology:
• Developed by ShengShu Technology, founded by researchers from Tsinghua University, as the successor to the Vidu Q1 and Q2 generations

Performance Characteristics:
• Native audio covers dialogue, sound effects, and music synchronized with the visuals
• Reference-to-video keeps subjects consistent across camera positions
‍
Prompting
Together AI API Access:
• Access Vidu Q3 via Together AI APIs using the endpoint vidu/vidu-q3
• Authenticate using your Together AI API key in request headers
• Create a video request from a text prompt, an input image, start and end frames, or up to four reference images, then poll until the video completes
• Available on Together AI with serverless access
‍
Applications & use cases
Short-Form Storytelling:
• Generate up to 16-second scenes with dialogue and music in one pass
• Keep characters consistent from reference images
• Bridge start and end frames into continuous shots

Advertising & Product Video:
• Turn product references into consistent ad clips
• Produce variants with synchronized sound design
• Animate product photos into short videos

Social & Creator Content:
• Create social clips with native audio from text prompts
• Animate images into narrative short videos
• Iterate on creative concepts quickly
‍

Model specifications

Model data

Model provider
Vidu
Type
Video
Main use cases
Text-to-Video
Image-to-Video
Resolution/Duration
1080p / up to 16s
Deployment
Serverless
Endpoint
vidu/vidu-q3
Price
$0.0975 / video
Input modalities
Text
Image
Output modalities
Video

Released
January 29, 2026
Category
Video

Run in Playground

Quickstart docs

Deploy Model

Vidu Q3

About model

API usage

Model card

Prompting

Applications & use cases