Any-to-Any Multimodal Intelligence aims to build a unified intelligent system capable of understanding, reasoning, and generating across arbitrary combinations of modalities. Unlike traditional multimodal models, which are typically limited to fixed input-output modes such as text-to-image or image-to-text, the any-to-any framework emphasizes flexible modeling capabilities in scenarios involving multiple inputs, multiple outputs, and interleaved modalities.

The core challenge lies in designing a unified architecture that supports multimodal collaborative modeling, in learning how to align semantics and control information flow in complex interleaved structures, and in evaluating generalization and scalability in real-world open environments. This direction therefore involves not only innovation in model architecture, but also the systematic construction of data organization methods, task design paradigms, and evaluation frameworks. This page serves as a central hub for our representative models, benchmarks, workshops, and surveys.