Meituan AI Model Lead Gu Qi Departs to Start Multimodal Venture
Gu Qi, a core member of Meituan's LongCat AI model team and head of its general agent unit, is leaving to found a startup focused on multimodal models, currently in talks with venture capital firms. He led the LongCat-Flash-Thinking-2601 model and contributed to LongCat-2.0, which Meituan released in late June with 1.6 trillion parameters and open-sourced in July.
Meituan's LongCat artificial intelligence large model team is undergoing a personnel change. Gu Qi, principal researcher on the LongCat team and head of the general agent unit, is set to depart, planning to start a business in the multimodal model field. He is currently in contact with venture capital institutions.
Gu Qi was a core contributor to the LongCat series of models and served as the lead for the LongCat-Flash-Thinking-2601 model. He graduated from the Department of Mathematical Sciences at Tsinghua University. Before joining Meituan, he worked as a researcher at ByteDance's Seed-LLM-Code team and at Ant Group.
Meituan officially released the LongCat-2.0 large model at the end of June, with total parameters reaching 1.6 trillion, and formally open-sourced it in July.