On September 9, 2025, Shengshu Technology officially launched the Q1 reference image generation function for its video large model Vidu, supporting input of up to 7 reference images to enhance and maintain subject consistency, and enabling the generation of multiple people in the same scene or across multiple scenes.