Kuaishou Technology has released an upgraded version of its Kling video-generation tool, focusing on reducing creation costs and time. The update aims to improve efficiency and accessibility for users. Kling 2.0 introduces enhanced features such as the Multi-Modal Visual Language (MVL) system, allowing seamless integration of text inputs with images or videos for precise control over character design and camera movements. Multimodal editing tools enable users to add, remove, or alter specific elements within generated videos, optimising workflow efficiency by iteratively refining videos without re-rendering entire scenes.
Visual quality enhancements include more realistic motion sequences and improved video fidelity. The platform can generate videos up to two minutes long at 1080p and 30fps. Sophisticated audio synchronisation, including built-in text-to-speech capabilities aligned with character lip movements, offers an end-to-end solution for professional-grade content. Kuaishou's Kling 2.0 integrates DeepSeek's language AI to enhance the creative process and improve prompt adherence, enabling the model to better understand and execute complex instructions. The global race in generative AI is accelerating, and Kuaishou aims to empower creators and lower the hurdles to producing sophisticated AI-generated visual content.
