Vidu: Vidu Q3 Turbo Image to Video API

Vidu Q3 Turbo Image-to-Video is an accelerated image-driven video generation model developed by Shengshu Technology for Q3, part of the Vidu Q3 series. Unlike the Text-to-Video variant in the same series, this model uses a user-provided static image as the first frame or key visual reference, combined with optional text descriptions, to generate videos with coherent dynamic content. The Turbo variant significantly enhances inference speed while maintaining high generation quality, making it ideal for image-to-video application scenarios sensitive to generation latency. Key capabilities include: Image-Driven Generation: Uses the input image as a visual anchor, strictly maintaining the consistency of visual characteristics such as character appearance, scene composition, and color style from the image, generating naturally coherent dynamic videos on this basis. High-Speed Inference: The Turbo version delivers significantly faster inference compared to standard and Pro variants, completing image-to-video tasks in considerably less time, suitable for real-time or near-real-time creative tool integration. Combined Image-Text Control: Supports simultaneous input of images and text prompts, allowing further specification of motion direction, action content, camera changes, and scene evolution through text descriptions, enabling fine-grained control over generated videos. High-Quality Motion Generation: Even under Turbo acceleration, generates smooth and natural motion effects including character actions, facial expression changes, object movements, and camera motions. Cost Efficiency: The Turbo version is more economical in inference resource consumption, suitable for large-scale batch image-to-video processing or high-concurrency call scenarios. Typical use cases include static image animation, e-commerce product showcase video generation, social media content creation, portrait animation, and dynamic interpretation of artworks.