Playground Guide
Follow these four steps to generate your first video. Each step takes under a minute.
For Seedance 2.0, any reference image or video that contains a human face must be uploaded to the asset library first, then used as a prepared asset in generation.
Open the Playground
Sign in and open the Playground. The top bar shows your current credits and a refresh button — keep an eye on it before kicking off long jobs.
Tip: if the balance is zero, generation requests will be rejected before they reach the model.
Pick a mode and output settings
Choose a generation mode and configure the output (aspect ratio, resolution, duration, output count, audio). Watch the short clip below for a tour of each option.
Upload reference assets
Two ways to bring images, videos, or audio into the Playground:
1. Direct upload from your computer
Drop or pick image, video, or audio files directly from your device. If the file contains a person's face, do not use direct upload — switch to Virtual Portrait below.
2. Virtual Portrait — for face assets you own
- Create an asset group first.
- Upload the matching images / videos / audio into that group.
- Wait a few seconds, then refresh the list.
- Once the asset indicator turns green, it's ready to use.
- We save your assets to your account, so you can keep reusing them later.
Write your prompt
Type your prompt in the input area. Once you've attached reference assets in Step 3, you can refer to them inside the prompt with their tags — for example image1, video1 — instead of describing them by name.
When the prompt looks right, hit Generate and wait for the result in the next step.
Helpful tips
Prompting patterns, common pitfalls, and ideas to get better results.