Meituan Open-Sources LongCat-2.5 Model: 1.6 Trillion Total Parameters, One-Million-Token Context Support
19 hour ago / Read about 0 minute
Author:小编   

On November 10, Meituan officially open-sourced LongCat-2.5 under the MIT License, permitting commercial inference deployment and secondary training without additional authorization. Designed for long-horizon agentic tasks, the model is natively multimodal and delivers significantly enhanced end-to-end performance in complex codebase development, full project construction, Visual Coding, and professional office scenarios. It features 1.6 trillion total parameters, activates approximately 48 billion parameters per token, and supports million-level context. For domestic AI chips constrained by memory and bandwidth, the model has undergone deep collaborative optimization across architecture, chip adaptation, and deployment strategies.