Soup.video
How to Make AI Videos with Seedance 2.0 on Soup.video (Tutorial)
Last updated on

How to Make AI Videos with Seedance 2.0 on Soup.video (Tutorial)


ByteDance Seedance 2.0 is a flexible multi-modal AI video model for text-to-video, image-to-video, and cinematic control with optional audio. This tutorial shows how to generate Seedance videos on Soup.video with clear pricing and prompt recipes.

Open the studio: Create with Seedance 2.0 →

Also available: Seedance 1.5 Pro for dialogue-heavy cinema clips.

Video editor timeline on a computer screen

Seedance 2.0 at a glance

  • Resolutions: 480p, 720p, 1080p
  • Durations: 4, 5, 8, 10, 12, 15 seconds
  • Aspect ratios including 9:16, 16:9, 1:1, 21:9, adaptive
  • Optional generate_audio

Seedance 2.0 pricing on Soup.video

Per-second rates (approx.):

ResolutionWith video/audio pathAlternate path
480p~$0.075/s~$0.10/s
720p~$0.15/s~$0.25/s
1080p~$0.36/s~$0.56/s

Example: 720p × 8s ≈ $1.20 on the standard video rate.
Draft on 480p to save while exploring prompts. Credits: Pricing.

For cheaper short dialogue tests, compare Seedance 1.5 Pro matrix pricing (from ~$0.04 for a 480p 4s silent clip).


Step-by-step: your first Seedance 2.0 video

1. Open /video/seedance2

2. Write a prompt with emotion + motion

Seedance handles human subjects and social energy well:

A dancer spins under purple stage lights, confetti in the air, energetic handheld camera, sharp costume detail, crowd silhouettes in background, 9:16

3. Choose resolution and duration

  • Ideation: 480p / 5s
  • Social publish: 720p / 8–10s
  • Hero/ad: 1080p / 8–12s

4. Toggle audio intentionally

Enable audio when you need ambience or speech-like energy. Disable it for silent B-roll to reduce cost.

5. Generate, review faces and limbs

If hands or faces drift, simplify the action and strengthen the subject description.


Image-to-video tips for Seedance

When uploading a reference:

  • Use a sharp subject with contrast from the background
  • Describe motion relative to the image, not a total scene rewrite
  • Keep wardrobe and identity cues consistent in the prompt (“same red jacket”, “same bob haircut”)

Seedance vs other Soup.video models

If you need…Prefer
Multi-shot director controlKling 3.0
Fast stylized social draftsGrok Imagine
Maximum photorealismVeo 3.1
Balanced cinematic + audioSeedance 2.0 / 1.5 Pro

API example

curl -X POST https://soup.video/api/v1/generations \
  -H "Authorization: Bearer sk_soup_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2",
    "inputs": {
      "prompt": "Cinematic close-up of espresso pouring into a ceramic cup, steam rising, slow push-in",
      "aspect_ratio": "9:16",
      "resolution": "720p",
      "duration": "8",
      "generate_audio": false
    }
  }'

Reference: Generate API.


Ready to cook? Start on Seedance 2.0 with pay-as-you-go credits — no separate ByteDance subscription required.