A3B (MoE, native multimodal)
MIT
Open-source native multimodal autoregressive model that unifies text, vision, and audio in a single discrete token framework under one training objective. Official release materials describe LongCat-Next as an industrial-strength multimodal foundation model built on a LongCat-Flash-Lite MoE backbone (A3B), with support for understanding and generation across modalities plus tool-calling workflows. The project emphasizes a unified discrete representation strategy (including DiNA-style native autoregression and discrete tokenizer/detokenizer interfaces), and reports strong broad multimodal benchmark competitiveness in release documentation. The model card and repo provide end-to-end usage examples spanning text chat, visual understanding/generation, audio tasks, and structured function/tool interactions, and the release includes open model weights together with tokenizer components to support further community research and development.