Infinigence AI Raises RMB 2.3 Billion to Expand Cross-Chip AI Infrastructure

Infinigence AI, a Chinese provider of AI infrastructure, has completed a RMB 2.3 billion Series C financing round, increasing its total funding to RMB 4.3 billion, as of 12 October 2026. The company develops software and systems enabling efficient operation of AI models across multiple types of processing chips, an important capability for AI deployment.

Cross-Chip Platform Supports Mixed-Processor AI Workloads

Infinigence AI’s platform integrates 37,000 petaflops of computing power distributed over 16 mainstream chip types. It supports mixed-chip training involving both Chinese and international processors, enabling greater flexibility in AI model development and deployment.

The company offers AI model inference services alongside infrastructure solutions tailored for enterprise AI agents, serving sectors such as gaming, healthcare, legal services and energy.

Efficiency Gains in AI Model Processing

According to Infinigence AI, its model-service platform’s daily token usage increased by 40 times between December 2025 and July 2026. The platform’s cross-cluster inference architecture is reported to have reduced the time to generate the first token by 51.5% and lowered the cost per token by 37.5%.

These improvements in inference speed and cost efficiency address key challenges in operational AI, supporting broader adoption and scalability of AI-based solutions in industrial and enterprise contexts.

Infinigence AI’s continued scaling and investment highlight the growing importance of versatile AI infrastructure capable of supporting heterogeneous processing environments, which is crucial for customised AI applications across the Asia-Pacific region.