ShengShu Technology unveiled an Artificial Intelligence solution for animated series production at SXSW 2026 in Austin, powered by its Vidu Q3 model. The company positioned the release as a response to a persistent gap in generative video workflows, where visually striking clips remain difficult to extend into consistent, serialized storytelling.
The system is built specifically for animation storytelling rather than general-purpose video generation. It targets several common production problems in Artificial Intelligence-generated animation, including non-human character instability, difficulty writing effective prompts, weak continuity across multiple shots, audio and visual performance misalignment, and limited reuse of intellectual property assets. Vidu says it improves character consistency through specialized training and enhanced multi-view stability, adds a prompt optimization assistant with structured visual parameters, and uses spatial structure control to preserve relationships and positioning between shots.
Vidu also emphasizes an audio-first generation pipeline, layered lip-sync processing, and improved audio-visual alignment to make dialogue feel more naturally connected to on-screen performance. For serialized projects, the platform includes a structured subject library designed for modular asset reuse, allowing creators to maintain recurring characters and visual elements across episodes. The broader goal is to support stable, scalable animation production workflows rather than one-off demonstrations.
Yihang Luo, CEO of ShengShu Technology, said Artificial Intelligence video generation has reached a stage where individual clips can look impressive, but producing consistent narrative content at scale remains difficult. He said the company aims to help creators move beyond isolated demonstrations and enable reliable production of serialized animated stories. Today, Vidu operates in more than 200 countries and regions, spanning industries such as animation, interactive entertainment, advertising, film, cultural tourism, gaming and more.
Vidu also said Vidu Q3’s Reference-to-Video feature will be launching soon. Originally released by Vidu in 2024, this feature supports multi-entity consistency with up to seven reference images to maintain stable character and scene continuity. Users can upload reference videos or images so the model can learn character movements, facial expressions, camera motion, and visual styles. ShengShu Technology was founded in March 2023 and describes itself as a multimodal Artificial Intelligence company focused on MaaS and SaaS tools for creative production.
