Model Introduction

Seedance 2.0 is positioned as a multi-modal controllable video model, but seedancemaker keeps the public web-app framing narrow and explicit.

Current Public Web-App Scope

• Text-to-video is available.
• Image-to-video is available with a single reference image path.
• Seedance 2.0 is treated as the paid tier.
• Public output options currently top out at 720p and 15 seconds.

Practical Interpretation

The model family may support broader multi-modal workflows in theory or in other environments. The public seedancemaker site focuses on the part that is clear, repeatable, and currently exposed to end users.

Core Inputs

• Text prompt: describe subject, action, framing, and motion.
• Reference image: use one still image when visual consistency matters.
• Model tier: choose between Seedance 1.5 Pro and Seedance 2.0 based on access and output needs.

Core Planning Rule

Plan prompts around the currently visible public limits first. If your concept depends on longer duration, higher resolution, or more elaborate input control, treat that as a separate research track instead of assuming the public generator already supports it.