Next‑gen image creation model unifying generation and editing in one architecture. It handles complex multimodal tasks (knowledge-based prompts, reasoning, and reference consistency) while delivering stunning 4K images with dramatically faster inference.