Seedance

Seedance 2.0 (BytePlus ModelArk) is a video model with native audio — it can generate a clip with sound in one pass.

#Get a key

Create a ModelArk key in the BytePlus console and paste it into ⚙ Settings → Seedance (or set ARK_API_KEY).

#Model & features

• Model id dreamina-seedance-2-0-260128, image-to-video and text-to-video, with native audio (generate_audio).
• First/last frame support (roled image inputs), plus reference image/video/audio roles.
• Watermark is always off.

#Pricing — read this before you buy credits

Seedance's billing has two sharp edges:

• Credit packs are model-bound. Seedance 2.0 credits don't cover the fast/mini tiers — buy the pack that matches the model you'll run.
• Rates vary by resolution and input class. Per-second cost climbs with resolution, the 4K rate is discounted, and reference-video input is ~40% cheaper per token. Brut's catalog derives a per-second USD estimate from the console rates:

The node shows the estimate before you run, and Film Shots itemizes it per clip.