Seedance 2.5 is live Limited-time offer: up to 51% off, from just $0.046/sec
Upgrade Now
ChatCut Docs

AI Generation

Compare Video Model Capabilities

Choose the model from the job you need to complete, then confirm its current inputs in Video Gen → Video settings before uploading references.

Use caseStart withWhy
Broad reference support, video editing, or extending a clipSeedance 2.5Accepts the widest mix of image, video, and audio references and exposes Edit and Extend modes.
General text, frame, or mixed-reference generationSeedance 2.0Full Seedance 2.0 generation path with up to 1080p output.
Faster or lower-cost iterationSeedance 2.0 Fast or miniSimilar input structure with lower-latency or lower-resolution tiers.
Image-led generation with several character or object referencesKling 3.0Combines frame slots and image references.
Generate from a few images or edit a short project videoGemini OmniSupports image-led generation and a separate short-video edit flow.

Model names alone do not guarantee the same duration, resolution, reference count, editing, or extension support. Changing models can remove incompatible fields and attachments; review the composer after every switch.

For the complete changing limits, use Model Capability Matrix. For the actual controls, use Video Model Selector.