Tripo AI 3D Generator
Mini toolsLongCat-2.0
By Meituan
Model family: LongCat
LongCat-2.0 is a large-scale Mixture-of-Experts language model with 1.6 trillion total parameters and approximately 48 billion activated per token. It introduces LongCat Sparse Attention (LSA), an evolution of sparse attention with streaming-aware, cross-layer, and hierarchical indexing enabling efficient native 1M-token context processing. The model also uses N-gram Embedding to expand the embedding space roughly 100x through n-gram token combinations. Pretrained on over 35 trillion tokens across 50K+ AI ASIC accelerators, LongCat-2.0 uses a multi-expert post-training pipeline with Agent Experts (coding, agentic tasks), Reasoning Experts (math, STEM, multi-hop reasoning), and Interaction Experts (instruction following, alignment). It integrates natively with agentic frameworks like Claude Code, OpenClaw, and Hermes, supporting repository-level code edits, automated task execution, and long-horizon agentic workflows.
New
Text
Gen 7
Overview
LongCat-2.0 is an open-source MoE language model with 1.6 trillion total parameters (~48B activated per token) and a 1M-token context window. It introduces LongCat Sparse Attention for efficient long-context processing and was pretrained on 35+ trillion tokens using AI ASIC hardware. Strong focus on coding, agentic, and reasoning tasks.
About Meituan
Chinese local services platform connecting consumers with food delivery, restaurant booking, travel, retail, grocery, and on-demand neighborhood services.
Industry:
Internet Marketplace Platforms
Company Size:
111298
Location:
Beijing, Chaoyang District, CN
Website:
meituan.com
Last updated: June 30, 2026
