🚀#Meituan releases LongCat-2.0 Set to be open-sourced 1.6T total params, ~48B active (33B–56B), 1M context window MOPD architecture integrating Agent, Reasoning & Interaction experts SWE-bench
Why CEREBRO kept it
LongCat 1M context window with agentic expert architecture
The text below is an automated extraction of the article at https://x.com/Chinazhidx/status/2071807454153961739, stored verbatim in the public cerebro-vault repository. Copyright remains with the original publisher (x.com).
Set to be open-sourced
1.6T total params, ~48B active (33B–56B), 1M context window
MOPD architecture integrating Agent, Reasoning & Interaction experts
SWE-bench
> LongCat 1M context window with agentic expert architecture
🚀#Meituan releases LongCat-2.0 Set to be open-sourced
1.6T total params, ~48B active (33B–56B), 1M context window
MOPD architecture integrating Agent, Reasoning & Interaction experts
SWE-bench Pro: 59.5 Outperforms Gemini 3.1 Pro (54.2), GPT-5.5 (58.6), Claude Opus 4.6 (57.3) https://t.co/XV6DhFFdyY
Backlinks
Appeared in 1 briefing