I'm LongbridgeAI. Here is the key information I extracted- The Mach M100 is a 5-nanometer automotive-grade AI chip delivering 1,280 TOPS of computing power per chip and utilizing a dynamic dataflow architecture.…

Key Takeaways
- Li Xiang said several robotics companies have asked about purchasing Mach M100 chips.
- Li Auto previously signed agreements for its chip business's first funding round, with plans to expand its external customer base and applications beyond cars.
Li Auto's (NASDAQ: LI) founder, chairman and CEO Li Xiang said the company will gradually supply its in-house developed Mach M100 chips to external customers, bringing technology already used in its vehicles to the broader embodied AI market.
Several robotics companies have asked whether they could purchase the Mach M100, Li wrote on Weibo on Tuesday, adding that computing infrastructure based on a dataflow architecture is important for embodied AI products.
"We appreciate these partners' trust and will gradually open up external supply of the Mach M100 chip," he said. The post did not disclose potential customers, a supply timeline or any signed purchase agreements.
The remarks further clarify the direction of Li Auto's chip business as it seeks to turn technology developed for its own vehicles into a business beyond car sales.
The Mach M100 uses a 5-nanometer automotive-grade process and delivers 1,280 TOPS of computing power per chip. Several new Li Auto models launched this year feature the chip as standard, and the company continues to improve its capabilities, Li said.
Li described the M100 as the world's first mass-produced AI chip with a dynamic dataflow architecture, saying the design aims to reduce the overhead associated with repeatedly moving data between computing units and memory in conventional architectures.
He also said a paper on the M100's architecture had been accepted into the Industry Track of the International Symposium on Computer Architecture (ISCA), making Li Auto the first automaker to have a paper accepted into that track.
The post coincided with a keynote speech by Li Auto CTO Xie Yan at TSMC's 2026 China Open Innovation Platform (OIP) Ecosystem Forum on the design and implementation of efficient AI inference systems.
Li Auto has already moved to bring external capital into its chip business. On October 16, the company announced agreements for its first equity financing round, totaling 1.625 billion yuan (about $242 million).
The financing is for Shanghai Mahe Xin Technology Co Ltd, known as MHX Technology. Li Auto committed 1 billion yuan, while external investors committed a combined 625 million yuan.
The arrangement also includes an option to subscribe for up to an additional 500 million yuan. Li Auto said at the time that it would retain control of the company and continue to consolidate its financial results after the financing, with the funds supporting chip development and commercialization.
MHX Technology plans to draw on a product portfolio spanning chips, domain controllers and operating systems to supply external customers while continuing to serve Li Auto vehicles, and to expand into applications beyond cars.
Li Auto's chip ambitions extend from assisted driving to data centers, report says
Li Auto is exploring an in-house cloud inference chip, and the project is still at an early stage.
($1 = 6.7132 yuan)
