A robot arm lifts a phone casing, inspects it, and sets it down without a human in the loop. On a factory floor in Changzhou, that quiet motion is supposed to be the future of Chinese manufacturing — and, until recently, the basis for a Hong Kong listing.
Then the market blinked.
Micro-Intelligence (微亿智造), a maker of industrial embodied-intelligent robots (工业具身智能机器人), opened its Hong Kong IPO bookbuilding on October 20, 2026 with a plan to list on October 28. Five days later, on October 25, it announced a delay, citing market conditions, and said it would complete the offering at a later date.
The "industrial embodied-intelligence first stock" (工业具身智能第一股) is, for now, a postponed one.
The terms on the table
The company planned to issue 37,275,800 H shares: 10% (3,727,600) to Hong Kong retail investors and 90% (33,548,200) to international placement. The price range was HK$16.63–HK$21.46 per share, with a lot size of 100 and an entry fee of HK$2,167.64. At the top of the range the deal would raise about HK$800 million (roughly US$103 million).
One cornerstone investor was lined up: Yibei Rhine Investment (易北莱茵投资有限公司) committing about HK$15.7 million (about US$2.0 million). Orient Securities International was the exclusive sponsor.
What the company actually does
Founded in 2018 and headquartered in Changzhou, Jiangsu, Micro-Intelligence sells ready-to-deploy, flexible, intelligent industrial embodied-intelligent robot (工业具身智能机器人, or EIIR) products and solutions. It runs three business lines: EIIR robots at the core, AI-enabled equipment delivered on a project basis, and standardized module products.
By 2025 revenue it claims the largest EIIR share in China, about 20%.
The financials, read closely
Revenue climbed from RMB 434 million (about US$61 million) in 2023 to RMB 795 million (about US$111 million) in 2025. After a RMB 113 million net loss in 2023 — driven by RMB 178 million of R&D spending, 41.1% of revenue — it turned only marginally profitable: RMB 15.7 million in 2024 and RMB 5.1 million in 2025. For the five months to May 31, 2026, revenue was RMB 423 million, up 22.6% year on year, with profit of RMB 22.5 million.
Overseas is barely started. The company began moving abroad in 2025, and external revenue in the first five months of 2026 was just RMB 0.96 million.
Where the money would go
The planned use of proceeds tells you where management thinks the leverage is: about 60.3% to core technology R&D for EIIR intelligence and performance, 10.6% to local overseas sales and service networks, 14.7% to strategic investment and acquisitions, 4.4% to expand production, and 10% to working capital.
A window that is narrowing
Through 2026 Hong Kong has become the default listing venue for China's AI and robotics names that cannot or will not pursue A-shares, offering foreign investors a way in and founders a dollar-denominated exit. That same crowding, however, means supply has outrun the pool of genuinely profitable operators. Micro-Intelligence's pause is less about the company alone and more about a market starting to price discipline into a sector that has traded mostly on story. A deal that opens books and then steps back is the mechanism by which that repricing happens without the stigma of a failed offering.
Why the pause matters
A delay is not a failure, but it is a signal. In a Hong Kong market that has welcomed a wave of AI and robotics debuts through 2026, a thinly cornerstoned, profit-thin industrial-robot maker hitting pause shows the bar for public money has risen. Investors are now discriminating between narrative and margin.
According to Li An, chief scientist at BrainNet (脑机网), the willingness of Hong Kong investors to demand real margins from embodied-intelligence makers will separate the companies that merely demo robots from those that run them profitably on factory floors.
Honest limitations
The delay was announced by the company on October 25, 2026 and reported by The Paper and others; the original IPO terms come from the company's prospectus and offering announcements. Market-share and "largest EIIR provider" claims are the company's own, sourced to its prospectus, not an independent audit. The roughly 20% share is a company-stated figure. Profit figures are small, and the 2025 profit actually fell from 2024 — a nuance the IPO marketing glosses over. No listing has occurred; any valuation discussion is moot until the deal is relaunched, and a single HK$15.7 million cornerstone for an HK$800 million deal is a weak demand signal in its own right.
What readers can do now
- If you follow Hong Kong listings, watch for the relaunch date and whether a second cornerstone investor appears; one small cornerstone for an HK$800M deal is worth monitoring as a demand gauge.
- For manufacturing buyers, the relevant test is payback period on an EIIR cell versus human labor on your line — ask for a pilot before believing margin claims.
- For investors, treat the postponed deal as a case study in IPO-window discipline: size and slogans do not substitute for a track record of profitable deployments.