Veo 3.1 is an AI video generation model from Google DeepMind, announced on October 15, 2025. It turns text or images into video and generates the audio in the same pass. Because dialogue, sound effects and ambient sound are made together with the picture, lips and timing line up.
Compared with Veo 3, it follows prompts more closely and produces better-looking video from images. A January 2026 update brought vertical 9:16 to videos made from reference images and added 1080p and 4K output.
Veo 3.1 Lite followed on March 31, 2026. Google describes it as its most cost-effective video model. Both Veo 3.1 and Veo 3.1 Lite are available on Seedance.kr, with the settings shown in the table below.
Based on the settings and credit rates currently on Seedance.kr.