2026 is shaping up as the first real IPO year for robotics companies, with more than 30 robotics firms seeking public listings.

Hundreds of cash-burning companies face elimination, and at least six are carrying tough valuation adjustment agreements.

Some say this is not a coronation, but a race for survival across the industry.

Whoever rings the bell first earns the right to stay alive; whoever is a step too slow may face industry elimination and a broken funding chain...

The Race for the Crown

On March 20, Unitree Robotics was accepted by the Shanghai Stock Exchange, with plans to raise 4.202 billion yuan.

Unitree’s ambition is to claim the crown as the first humanoid robotics stock on China’s A-share market.

The news swept through the industry. Unitree is, without question, a leading contender for that crown.

The expected listing window is June to July, and the industry puts its odds of taking the crown as high as 90%.

In second place is Leju Robotics.

On October 30, 2025, Leju began the STAR Market tutoring filing process. It is expected to complete tutoring between March and June 2026, and is likely to become the second humanoid robotics company on the STAR Market.

The third is Deep Robotics.

In December 2025, Deep Robotics began its STAR Market tutoring process and is expected to complete it between April and June 2026, putting it behind Leju. The fourth is Tianlian Robotics.

It filed for tutoring on July 30, 2025, and is expected to complete the process in 2026. With four companies racing at the same time, the entire industry is watching closely to see who ultimately wins.

How much is that label worth? UBTech offers the clearest example.

In December 2023, it listed in Hong Kong as the first humanoid robotics stock.

Although UBTech accumulated more than 5.4 billion yuan in losses from 2020 through the first half of 2025, the industry voice and resource concentration that came with being the first stock cannot be measured by a financial report alone.

Whoever lists first becomes the origin point of the industry’s coordinate system. Every later entrant has to tell its story in relation to that benchmark.

Hong Kong stocks are far less liquid than A-shares, which makes the crown of the first A-share humanoid robotics stock much more valuable.

The industry calls it the battle for the crown.

Who will win this fight? For now, Unitree has taken the lead. After all, its IPO application has already been accepted.

The other three, based on public information, are still in the tutoring stage.

Lining Up to Go Public

Beyond these four companies fighting for the first crown, several others are queued up behind them, crowding into the A-share market.

They include robotics companies such as Songyan Power, Xinghaitu, EngineAI Robotics, and MagicLab.

There are also many component players, including Zhitong Technology, Xinjian Transmission, and KinLink Transmission.

Outside A-shares, Hong Kong is another popular path to survival.

The most closely watched Hong Kong rumor concerns AgiBot. Although the company says it has no clear plan for a Hong Kong IPO for now, market chatter says it may file in the first quarter of 2026 and target a third-quarter listing.

Another is Galbot, which completed its shareholding reform in 2025 and is said to be preparing for a Hong Kong listing.

In addition, about 20 robotics companies have already filed with the Hong Kong Stock Exchange.

Of course, many robotics companies are pursuing both paths, keeping A-shares and Hong Kong in play at the same time.

AgiBot, for example, is also preparing for the STAR Market.

Why are more than 30 robotics companies determined to squeeze into the market in 2026?

Capital Pressure

For many robotics companies, going public is the only way out.

First, investors need an exit.

China saw a peak in fund formation in 2015 and 2016. Based on fund lives of seven to 10 years, that batch is entering liquidation in 2025 and 2026.

They are in a hurry to get their money back.

So far, at least six robotics companies have signed valuation adjustment agreements with investment institutions.

They include Unitree, Hai Robotics, Geek+, and others. Unitree’s agreement has been described as the toughest, harshest, and most representative in China’s robotics circle.

Investors put money into Wang Xingxing’s company, with three terms written in black and white: first, it must successfully list on the STAR Market before December 31, 2026.

Second, its market value at listing must not be lower than 12 billion yuan.

Third, its full-year 2026 revenue must exceed 3 billion yuan, with net profit above 300 million yuan.

Breaking any one of those terms counts as a default. If that happens, Wang Xingxing personally must pay to buy back all investor shares.

It is not just the principal. He would also have to pay 15% interest. A simple calculation: the Series C round raised about 700 million yuan, and with interest, Wang would need to come up with more than 800 million yuan in cash in one shot.

But before the listing, the agreement was terminated because many of its strict terms crossed red lines for going public.

Companies sign these agreements to raise money and terminate them to list in compliance. For many firms, an IPO is not just a financing tool, but the only exit for unwinding valuation adjustment agreements and easing pressure on the funding chain.

Second, money in the primary market has become harder to secure. More than 140 companies are still burning cash. The capital market remains hot, but valuations have already priced in expectations for the next several years.

In March 2025, GSR Ventures’ Zhu Xiaohu publicly announced that he was exiting humanoid robotics projects in batches, while many professionals also called for the industry to cool down.

Since 2025, institutions including Sequoia China and IDG have mostly stopped launching new early-stage humanoid robotics investments, following on only in existing leading portfolio companies. Xinding Capital predicts that more than 100 companies raised financing in 2025, but by 2026, only 10 to 20 may remain.

“Honestly, the basic reality is that robotics projects before Series B are all finding it very difficult to raise money. The window is already gone,” said the co-founder of one robotics company.

Valuations are also shrinking fast. “Investment institutions keep pushing valuations down. We wanted a 3 billion yuan valuation for our Series B, but investors pushed it down to 1 billion yuan,” the co-founder said.

Third, state capital needs exits, and policy has opened a window.

In September 2024, the Hong Kong Stock Exchange lowered the threshold for Chapter 18C listings. On June 18, 2025, the China Securities Regulatory Commission reopened the STAR Market listing channel for unprofitable companies, and ChiNext followed suit.

It is rare for all three channels to open at the same time. Once exit routes became clear, state capital moved in immediately. IT Juzi data show that among the top 20 robotics companies by financing amount in the third quarter of 2025, 15 had state-backed capital behind them.

Xiongan Fund invested in Mech-Mind Robotics, Hefei backed EngineAI, and the Beijing Robotics Industry Development Investment Fund placed bets on Songyan Power and Deep Robotics.

The logic is simple: money dares to come in only when there is an exit path, and once the money arrives, companies have the confidence to sprint toward IPOs.

State capital wants to attract investment, old investors want a dignified exit, and companies need lifeline funding for R&D. The three sides found common cause, producing an IPO wave in which more than 30 companies are racing at once.

But ringing the bell does not mean reaching shore.

Who Pays?

More than 30 companies are pushing frantically into the stock market, but the results they have delivered so far are not reassuring.

In 2025, China’s embodied intelligence sector raised 73.5 billion yuan. But the combined revenue of humanoid robotics companies was still below 10 billion yuan.

Can revenue like that support these valuations? Even for companies with orders in hand, the details are worrying.

First, many of the buyers are universities and research institutions.

In the first nine months of 2025, 73.6% of Unitree’s humanoid robots were sold to universities and research institutions. Revenue from robots actually working in factories was just 15.7 million yuan, less than 1% of total revenue. Investor Chen Yuan noted: “Universities do not need the robots to be smart. They buy them to train algorithms. But factories need them to do work, and at the current level of intelligence, it is hard for procurement managers to sign off.”

And demand from laboratories nationwide has a ceiling. Unitree knows this too.

That is why 2.022 billion yuan of its 4.2 billion yuan fundraising plan is earmarked for AI large models, preparing for the move “from the lab to the assembly line.” Whether Unitree’s 42 billion yuan valuation can hold will depend on how fast its robots can enter factories and do real work.

Second, many of the buyers are companies within the same circle.

Leju, Galbot, Xinghaitu, Qianxun Intelligence, and Zibianliang Robotics are all valued at more than 10 billion yuan, but they remain some distance from real profitability.

Looking through the customer lists of this group of robotics companies, the biggest buyers are often their own shareholders. In September 2024, Galbot and Meituan, one of its angel-round backers, deployed a retail terminal project.

In March 2025, UBTech sold 500 robots to early shareholder Easyhome. In June 2025, AgiBot and Hill Robotics, a subsidiary of shareholder Wolong Electric Drive, jointly won a 24.74 million yuan project. In July 2025, Leju signed a strategic cooperation agreement with shareholder Dongfang Precision.

Chen Yuan is not surprised by this: “The technology is not mature enough for unfamiliar customers to willingly pay. First, you get insiders to buy and push up shipments. Once the data looks better, you can keep raising money and turning over capital.”

Will the secondary market really be willing to take over from here?

The rush by more than 30 companies to list together may not be a sign that the industry is maturing. It may be the final act of a bubble the primary market can no longer support.

However compelling the story, real orders paid in real money will ultimately decide the outcome.

How long can a story built purely on capital keep going?