LongCat-2.0 is a large language model released by Meituan, released under the mit license, with a claimed knowledge cutoff of 2026-07-20. Its weights, configuration, and tokenizer files could not be hashed successfully -- see the error below by Weight Registry.
Description
LongCat 2.0 is a 1.6-trillion-parameter Mixture-of-Experts model from Meituan, released June 30, 2026 under MIT license with roughly 48B parameters active per token. It's built for agentic coding and repository-level work using a custom LongCat Sparse Attention mechanism aimed at 1M-token context, and Meituan reports it was trained entirely on domestic (non-Nvidia, non-AMD) compute hardware.
Metadata
Hugging Face stats
1,773 downloads on Hugging Face · 273 likes · pipeline: text-generation · library: LongCat-2.0
Weight Hashes
Hashing failed -- see the note below for the error.
Error
('Connection broken: IncompleteRead(3378231148 bytes read, 15244549148 more expected)', IncompleteRead(3378231148 bytes read, 15244549148 more expected))