LongCat-2.0 from @Meituan_LongCat is free in Nous Portal for one week. It is a 1
Source
portal.nousresearch.com
Author
NousResearch
Date
Context
Nous Research's one-week Portal promotion offered access to LongCat-2.0, a coding-oriented model from Meituan. Its size needs two separate numbers: the model documentation lists 1.6 trillion total parameters, with about 48 billion activated per tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition →. The mixture-of-expertsA model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.Full definition → design uses a portion of the model's parameters at each step.
Nous's post also describes a 1M-token context windowThe maximum amount of text a model can consider at once — its working memory for the current conversation or task.Full definition →, and the documentation says the model was trained on long-context data. A long context matters for work spanning a repository, but whether a particular repository fits depends on its size. The promotion offered a way to try the model; its one-week term did not establish permanent free access.
Terms in this piece · Glossary
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
mixture-of-experts — A model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.
context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
