DreamActor V2

DreamActor V2 is a video generation model developed by ByteDance and made available through the WaveSpeed provider. It takes two inputs — a base image and a motion reference video — and generates a video in which the subject from the image performs the movements captured in the reference footage. This motion transfer approach allows users to animate still images without manually keyframing or scripting movement. The model is priced at $0.05 per second of generated video and supports a context window of 1,000 tokens.

DreamActor V2 is well suited for use cases that require animating portraits, characters, or other subjects using real-world motion as a guide. Because the model accepts a video URL as the motion source, creators can supply existing footage to drive the animation rather than relying on text prompts alone. This makes it particularly useful for applications in digital content creation, virtual avatars, and motion-driven storytelling. The model was added to MindStudio in February 2026 and is currently available for use.