Vidu Q3

• 🎬 Shengshu's image-to-video model in the Vidu Q-series.
• 🆕 The image-conditioned variant of the Vidu Q3 generation.
• 📅 Released January 2026.
• 🎨 Tuned for anime and stylized video output.
• ⚡ Positioned as an efficient video generation system.
• 💬 Pairs with the [[sibling:vidu-q3-text-to-video|Vidu Q3]] text-to-video model.
• 🏢 Built by Shengshu Technology (生数科技), a Chinese AI company.

Shengshu

Shengshu is an AI research company focused on generative video technology. The lab is behind the Vidu family of models, which target high-quality video synthesis from both text and image inputs. Their work positions them among a growing cohort of…

Vidu Q3 Image-to-Video is the image-conditioned member of Shengshu Technology's (生数科技) Vidu Q3 video generation family. Rather than starting from text alone, it takes a static input image — optionally guided by a text prompt — and animates it into a short moving clip. It is part of the broader Vidu suite of generative video tools developed by Shengshu, a Chinese AI company.

Within the current generation, the model sits alongside its sibling Vidu Q3 text-to-video system, which generates clips directly from written prompts. The two share the same Q3 branding and release window, offering complementary entry points — image-driven versus prompt-driven — into the same family of video generation capabilities.

According to its catalog positioning, Vidu emphasizes efficient video generation with particularly strong support for anime and stylized output, making the image-to-video variant a fit for animating illustrations, character art, and other non-photoreal source images into motion.

Detailed independently verified specifications and benchmark results for this model are not available from primary or top-evaluator sources at the time of writing, so this description is limited to its confirmed role: an image-to-video model in Shengshu's Vidu Q3 lineup, released in January 2026.