Meituan launches LongCat-2.5, a native multimodal model for long-horizon agents

Meituan released LongCat-2.5 on November 2nd, a native multimodal model built for long-horizon agentic work with 1.6 trillion total parameters and approximately 48 billion activated per token, according to Meituan's announcement. The model adds native vision to long-horizon coding and ships with a 1M-token API window, with GitHub and Hugging Face releases still to come. Meituan, co-founded by Wang Xing, is pushing frontier-scale agentic coding with low cached-input pricing.

Meituan launches LongCat-2.5, a native multimodal model for long-horizon agents Meituan's 1.6-trillion-parameter model adds native vision to long-horizon coding, with a 1M-token API window and GitHub and Hugging Face releases still to come. By RuntimeWire Staff https://runtimewire.com/author/runtimewire-staff · Published · Updated Primary source: Hacker News https://longcat.ai/blog/longcat-2.5/ Why it matters LongCat-2.5 shows a large platform company pushing frontier-scale agentic coding with native vision, a 1M-token window, and low cached-input pricing. Meituan, co-founded by Wang Xing, released LongCat-2.5 https://longcat.ai/blog/longcat-2.5/?ref=runtimewire on November 2nd, according to Meituan's announcement https://www.meituan.com/news/NN261102280001343?ref=runtimewire . The native multimodal model is built for long-horizon agentic work, with 1.6 trillion total parameters and approximately 48 billion activated per token. The release pushes Meituan's model line further into sustained coding work, where an agent has to hold a project together across many steps.… When you purchase through links in our articles, we may earn a small commission https://runtimewire.com/affiliate-monetization-standards . This doesn’t affect our editorial independence.