Kling AI Launches 3.0 Model, Ushering in an Era Where Everyone Can Be a Director
Kling AI launched its 3.0 model series, featuring Video 3.0 and Image 3.0, with major upgrades in consistency, photorealism, and native audio generation.Summary
Kling AI has announced the release of its Kling 3.0 model series, which includes Video 3.0, Video 3.0 Omni, Image 3.0, and Image 3.0 Omni. These models are built on an integrated unified training framework supporting full multimodal input and output (text, images, audio, video) within a single workflow, allowing for better narrative control and prompt adherence.
Key enhancements in Video 3.0 include improved element consistency via reference uploads, native audio generation in multiple languages and accents (English, Chinese, Japanese, Korean, Spanish), extended video duration up to 15 seconds, intelligent multi-shot storytelling, and photorealistic output. Video 3.0 Omni adds advanced storyboarding features for precise shot specification. The Image 3.0 models now support 2K and 4K ultra-high-definition output.
The Kling 3.0 lineup embodies the Multi-modal Visual Language (MVL) framework, marking a shift toward sophisticated professional orchestration. The models are currently available for early access to Ultra subscribers and will soon be public. Since its launch in June 2024, Kling AI has served over 60 million creators and is seeing adoption across the film and advertising industries.
(Source：Prnewswire)