Beijing-based Approaching.AI, also known as Qujing Technology, has raised more than RMB 1.3 billion, about $194.66 million, in Series B financing. The round is one of the largest recent bets on a problem every AI company now faces: the rising cost of running models at scale.
The investors include Zhejiang Finance Capital, Xiaoshan Capital, ICBC Investment, Zhongguancun Science City, Xinglian Capital, Qianjiang Century City, SenseTime, and Wangsu Science & Technology. Total funding reported for the company during 2026 now stands at approximately RMB 2.4 billion, while a valuation for this round was not disclosed.
Approaching.AI focuses on inference, the stage where a trained model generates responses. As assistants and agents make more model calls to complete a single task, inference becomes a major operating expense. The company builds software and computing infrastructure designed to improve computing utilization, memory management, and token-generation efficiency.
It also aims to improve performance across different processor architectures, including Chinese AI hardware. That breadth matters as companies look for ways to keep response times fast without being locked into a single hardware path.
The funding reflects a shift in investor attention from building ever-larger models to making existing models affordable to operate. Efficient inference helps open-source and commercial models alike, which is why hubs such as Hugging Face and local-running tools such as Ollama have become important parts of the wider ecosystem for developers testing and deploying models.
Still, the real test will be commercial evidence: comparable cost, performance, and reliability under real workloads. If Approaching.AI can prove those gains outside controlled settings, cheaper inference could lower the barrier for smaller teams to build AI products that stay affordable as they grow.
Source: TechStartups
