According to monitoring by Dongcha Beating, AI agent developer Mindverse has announced the completion of a Series A financing round led by Meituan, with participation from Yuanhe Puhua, Shokz, Variable Capital, and existing shareholders, totaling nearly $50 million. Gao Hu Capital acted as the exclusive financial advisor. Mindverse was co-founded by Chen Kaijie and Andrew (who previously collaborated with Yao Shunyu on the ReAct/FireAct paper), with a core R&D team from DeepSeek, ByteDance Seed, xAI, etc., dedicated to building native agent models through post-training and continuous learning technologies. Unlike conventional methods that rely on external prompt concatenation, Mindverse firmly believes that the core capabilities of agents stem from the model's 'post-training.' The company uses LoRA as the technical foundation for continuous learning and parameterized memory, achieving lossless LoRA reinforcement learning for a trillion-parameter MoE model by the end of 2025, reducing post-training costs to one-tenth of full parameter training. Through the Mixture of LoRA architecture, thousands of independent 'skill packages' (LoRA) representing different preferences or business experiences can be mounted on a shared base, supporting dynamic rapid activation (within seconds) and clean privacy isolation. Mindverse announced it will open source its latest trained 750B parameter agent model, which includes a 744B pre-trained GLM 5.1 base and 6B of LoRA, making it the world's first result of reinforcement learning post-training on GLM 5.1, achieving SOTA on four major benchmarks: Living Bench, Vita Bench, A2UI, and PinchBench. On the product side, Mindverse employs a bidirectional flywheel that iterates collaboratively between model training and its consumer-facing life agent application Macaron (2 million users, 100,000 daily active users); commercially, it collaborates deeply with Huawei Cloud, Microsoft Cloud, and others to open up low-cost, high-concurrency LoRA training and inference infrastructure to enterprises.