Seedance 2.0 on Genso AI: Reference Video, Audio & Multi-Image Cinematic Video

Seedance 2.0 is ByteDance's latest cinematic video line available inside Genso AI's Video Lab. It targets creators who need references — not only text — to lock identity, camera rhythm, or audio cadence.

Outside Genso AI, Seedance family models are often discussed alongside Dreamina / CapCut ecosystem rollouts; Genso AI is a studio-first path: one credit wallet, same history grid as your Kling and Veo tests.

Prompt — describe subject, action, camera, and style.

First frame / last frame — optional stills for start and end composition. When either first or last frame is set, reference video and reference audio are unavailable (mutually exclusive with those frame plates in Genso AI).

Reference images — up to seven stills for identity or wardrobe continuity.

Reference video — 2–15 seconds for style or motion transfer when you are not using first or last frame stills.

Reference audio — up to 15 seconds; requires a reference image or video per Genso AI's validation rules.

Duration — 5–15 seconds slider.

Resolution — 480p or 720p.

Generate Audio and Camera Fixed toggles round out the control surface.

Pre-plan exclusivity: Decide up front whether you are anchoring with a first or last frame still or with reference video — switching mid-project wastes iterations.

Trim references tight: Shorter, cleaner reference clips reduce ambiguity and usually spend fewer credits for the same creative intent.

Match Seedance 1.5 Pro: If you only need expressive dance motion without the 2.0 reference stack, Seedance 1.5 Pro remains in the model list for simpler jobs.

Third-party articles sometimes cite resolutions or reference counts that differ by region or provider. Genso AI's Video Lab shows the authoritative limits for Seedance 2.0 (duration slider, resolution select, and upload hints). Use those labels when planning a client deliverable.