The Model
Seedance 2.0 Mini (upstream API identifier seedance-2.0-mini, URL slug seedance-2-0-mini) is a lightweight video generation model from ByteDance Seed, the research division behind Doubao, Dreamina (Jimeng AI), and CapCut’s AI pipeline. It shares the same unified audio-visual diffusion transformer architecture as the standard Seedance 2.0 and the flagship Seedance 2.5, predicting pixel motion, camera parallax, lip-synced dialogue, and environmental foley in a single joint forward pass.
Where the standard Seedance 2.0 scales up to native 4K and Seedance 2.5 targets multi-shot narrative continuity at 1080p, Mini is built to be cheap and fast. It caps at 480p and 720p, trades rendering headroom for rapid turnaround, and costs roughly half the per-second rate of its standard sibling. The design intent is explicit: use Mini for drafts, prompt tests, hook variants, and high-volume social batches — then promote only the winning take to a higher-resolution model for the final render.
On this studio, Seedance 2.0 Mini runs in the same model picker as Seedance 2.5, Seedance 2.0, Seedance 2.0 Fast, Wan 3.0, and MiniMax H3, all drawing from a single shared credit balance. No separate account, no separate API key, no separate billing.