MiniMax

Stepfun, founded in 2023 in Shanghai, China, is an AI large language model application service provider. The company has launched two consumer products — Yuewenyuewen and Maopao Duck — focused on multimodal output, and has partnered with OPPO and Honor in the enterprise market, aiming to become a mobile-side AI assistant.

ModelBest (面壁智能), founded in 2022 in Beijing, China, is an AI large language model company. The company has not disclosed specific funding amounts or valuation.

01.AI (零一万物), founded in 2023 in Beijing, China, is an AI large language model and AIGC service provider founded by Kai-Fu Lee. The company released its first open-source large model 'Yi', with Yi-34B ranking first on the English pre-trained open-source model leaderboard and first among publicly available large models on the third-party C-Eval leaderboard in Chinese capability.

Baichuan AI, founded in 2023 in Beijing, is one of China's six leading large language model startups, focused on generative AI and LLMs with applications in finance, education, and healthcare.

Shenzhen Mindverse Technology (Mindverse), founded Oct 2023, is a globally-oriented AI company focused on post-training and continuous learning for agent models. Using reinforcement learning plus LoRA on top of general foundation models, it aims to turn models from 'knowing a lot' into 'getting things done' with low-cost personalization. Its consumer product Macaron, billed as the world's first personal agent, instantly turns conversations into personalized mini-apps; it has launched in North America, Japan and Korea with 2M+ users and ~100K DAU. On the tech side it open-sourced Macaron-V1-Preview, a 749B-parameter agent model based on GLM 5.1 (202K context, MIT license). Its ~20-person core R&D team draws from DeepSeek, ByteDance Seed and xAI, with Tsinghua/MIT/Duke backgrounds.

Wuxian Guangnian (Infinite Light Year), founded in 2022 in Shanghai, China, is a developer of AI large language model technologies.