LongCat27B Activation

Open source LongCat-Flash-Omni: 128K context + text/image/audio/video to speech output

I. Abstract LongCat-Flash-Omni is an open-source, multimodal (Omni-modal) model from Meituan's LongCat team. It extends the ScMoE architecture of Long...