Code & Chain · Signal Desk

Meituan Open-Sources LongCat-2.0: 1.6 Trillion Parameter Model Targeting Agentic Coding and Long Context

Original sourceaitoolly.com

Summary

Meituan officially open-sourced LongCat-2.0, an agentic coding model with 1.6 trillion total parameters. Using sparse attention and N-gram embedding, its average active parameters are only about 48 billion, enabling extremely long code context handling. The model focuses on code understanding, gene…

Key points

  • Offers developers a powerful open-source, locally deployable model choice for handling large codebases
  • The emergence of large-scale open-source coding models may prevent advanced agentic development tools from being monopolized by a few closed models
  • Software companies can use it for internal codebase analysis, refactoring, or automated testing without API call costs or data leakage concerns

Editorial note

This page is Code & Chain's editorial summary of public sources. It may be prepared with AI assistance and published through an automated workflow. Refer to the original sources; this content is not investment, legal, or tax advice.