According to The Decoder, Meituan's LongCat team released LongCat-2.5 model weights under MIT license on November 15. The mixture-of-experts (MoE) model features 1.6 trillion total parameters with approximately 48 billion activated per token, supporting a 1 million token context window. Training was conducted entirely on Chinese hardware.

The model scored 88.4 on Terminal Bench v2.1, compared to Claude Opus 5's 89.1 and GPT-5.6 Sol's 88.8. Meituan also released a 4-bit quantized version with self-hosting guidance based on 194 Huawei Ascend chips, without providing Nvidia GPU options.

View Source

Disclaimer: This content is for informational purposes only and does not constitute investment advice or encourage the buying or selling of crypto assets. It is intended to provide information about industry developments and is not directly related to the services offered by Gate TR. Although information obtained from third-party sources has been carefully reviewed, Gate TR assumes no responsibility for any errors or omissions. Crypto assets involve high risks and may experience significant price fluctuations. Before making any investment decisions, you should assess your financial situation and risk tolerance, conduct your own research, and make decisions independently.