AiShi Technology has published a technical report detailing its latest innovation—the real-time, full-modal world model, PixVerse R2. Building on the foundation laid by its predecessor, PixVerse R1, this new model delves deeper into the realm of real-time world simulation. It employs a unified Omni Causal AR architecture, which is adept at handling multimodal inputs. This enables PixVerse R2 to seamlessly simulate a world where generation, interaction, and evolution unfold concurrently, mirroring the dynamic nature of the real world and aligning with the nuanced expectations of English-speaking audiences familiar with cutting-edge technological advancements.