Skywork AI Unveils Open-Source Matrix-Game 2.0 for Real-Time Interactive Generation

Skywork AI has launched Matrix-Game 2.0, an open-source interactive world model offering real-time, long-sequence generation, as announced in a press release. This release marks the first open-source solution in the industry for real-time, long-sequence, interactive generation in general scenarios.

Matrix-Game 2.0 is designed to enhance user interaction by providing high-frame-rate, real-time generation capabilities. It supports seamless video generation at 25 FPS, allowing for minute-long interactive sequences. The model also boasts cross-scenario generalization, enabling it to adapt to various environments and visual styles, from urban landscapes to oil-painting aesthetics.

In addition to its real-time capabilities, Matrix-Game 2.0 emphasizes physical consistency, ensuring that generated characters exhibit realistic movements across complex terrains. This feature enhances the immersion and controllability of virtual environments, making it suitable for applications such as virtual humans, game engines, and embodied AI.

Skywork AI's open-source initiative aims to foster collaboration with global developers and users, paving the way for advancements in artificial general intelligence (AGI).