Meituan releases LongCat-2.0, a 1.6T-parameter open MoE model with 1M context
Meituan Releases LongCat-2.0: A 1.6T-Parameter Open MoE Model with Native 1M Context and LongCat Sparse Attention
MarkTechPost
Summary
Meituan released LongCat-2.0, a 1.6 trillion-parameter Mixture-of-Experts model that activates approximately 48 billion parameters per token. It features native 1-million-token context built on LongCat Sparse Attention.