What Anthropic's IPO Reveals About How Chinese Large-Model Companies Are Valued
Using Anthropic's valuation fluctuations following its IPO as a benchmark, the article analyzes how the valuation logic for Chinese large-model companies is shifting toward their ability to deliver commercial results. A report by the China Academy of Information and Communications Technology shows that the normalized average price of domestic model services rose from 6.8 yuan in the second half of 2025 to 13.46 yuan so far this year; the author argues that whether companies continue using APIs after price increases will become a key measure of models' pricing power.
Why read this
SelectedAI score 84
Drawing on Anthropic's valuation fluctuations following its IPO and price increases for domestic model services, the article examines how the ability to deliver commercial results is becoming the basis for valuing large-model companies.
Topics
Body · AI translation
What Anthropic’s Bell-Ringing Tells Us About Pricing China’s Large Models — Light Cone Intelligence. Download the app for exclusive early access. November 20, 2026, 08:52:01, from Sichuan. Updated November 20, 2026, 09:05:33
Anthropic’s IPO is more like a stress test for the entire AI industry.
By Wei Linhua
Edited by Liu Junhong
At 11:48 a.m. Eastern Time on November 19, Anthropic (ticker: ANT), hailed as the “largest IPO in human history,” officially began trading on Nasdaq.
Compared with Anthropic’s final offering price of $83, the stock opened sharply higher at $111.50, a surge of 34.3%, and at one point climbed to $115.10 intraday. On a fully diluted basis, Anthropic’s valuation peaked at $2.02 trillion, surpassing SpaceX, which went public in June this year.
Yet the post-opening euphoria lasted less than half an hour. By the close, Anthropic’s market capitalization had been on a roller-coaster ride.
Anthropic’s shares began plunging in the afternoon, falling as low as $92.72 intraday before closing at $97.17. Although that was still a gain of 17.07% from the $83 offering price, Anthropic had shed approximately $324.4 billion in market value from its intraday high by the close. On a fully diluted basis, the company’s valuation fell from $2 trillion to around $1.76 trillion.
As the bell rang across the ocean, nearly every large-model founder, investor, and secondary-market analyst in China’s AI sector stayed up late watching the market. They were not focused solely on the rise and fall of Anthropic’s stock, but searching for an answer: Now that the “anchor” for global AI valuations has dropped into the deep waters of the public market, how exactly should Chinese AI models be priced next?
After all, on the eve of the bell-ringing, Anthropic’s valuation had repeatedly declined amid debate over its pricing—from the initial $2 trillion valuation assigned by investment institutions, to the valuation cut during Anthropic’s Q3 roadshow, and then the impact of the U.S. midterm elections and Nvidia’s sudden leadership change.
For Anthropic, pressing ahead with an IPO at this juncture is more like a stress test for the entire AI industry.
Anthropic’s IPO has never been a story about whether it is “expensive.” It is a story about how its value should be calculated.
And the new yardstick Wall Street is using to scrutinize Anthropic today will be applied unchanged to every Chinese large-model company.
Anthropic’s Valuation Rose
But the Excess Had Already Been Squeezed Out
As a “bellwether” for large-model companies worldwide, the 17% rise in Anthropic’s share price is entirely normal.
Investor appetite remains strong. According to information reported by Bloomberg, subscription orders amounted to approximately six to seven times the number of shares available, driven primarily by large indications of interest from institutional investors.
As a result, Anthropic not only had the confidence to raise its price but also planned to capitalize on favorable market conditions to secure even more funding.
By November 18, Anthropic had ultimately raised its share price from the $70–$78 range to $83. It also forcibly expanded the number of shares offered by 19%, from the originally planned 825 million to 980 million, raising nearly $20 billion in additional cash.
With demand for the shares outstripping supply, Anthropic also had a host of computing-power suppliers backing it:
Just before the listing, a group of technology giants that could be called a “global computing-power all-star team”—including Nvidia, Google, Microsoft, AMD, Broadcom, Micron, and partners of TSMC—banded together and agreed to purchase $26 billion worth of Anthropic shares. Google and Nvidia were the biggest spenders, each putting up $10 billion.
By stepping forward, these computing-power suppliers and cloud vendors effectively tied themselves firmly to Anthropic while securing computing-power orders in advance.
The $26 billion investment is more like a “strategic lock-in fee” paid by the giants. As long as Anthropic uses the money to continue its aggressive expansion and train its next-generation models, the funds will ultimately come full circle and turn into orders, flowing through procurement contracts back into Nvidia’s chips, Broadcom’s servers, and the data centers of major cloud providers.
So who are Anthropic’s largest shareholders? The answer is Amazon and Google. According to data disclosed by Anthropic, Amazon holds a total stake of 14.7%, followed closely by Google with 13.7%.
Although the company’s market capitalization ultimately settled at around $1.76 trillion, viewed over a longer timeline, the seeds of today’s fluctuations in Anthropic’s market value were actually sown three months before its listing.
Back in August, the market was still enthusiastically calculating a $2 trillion valuation for Anthropic. The pricing logic at the time was simple and blunt, mirroring China’s private market frenzy over large models: As long as the models kept improving and revenue growth continued to soar, there was no ceiling on the company’s valuation potential.
But reality poured cold water on that enthusiasm. Beginning in October, Anthropic experienced several “moments” of falling valuation:
First, revenue growth fell short of expectations, prompting investment institutions to reassess its value. During its October roadshow, Anthropic disclosed that its quarter-on-quarter revenue growth had fallen from 144% in Q2 to 47% in Q3. Investment institutions consequently took some of the froth out of its valuation, marking it down to $1.5 trillion.
That adjustment still reflected doubts about Anthropic itself. The subsequent decline in the company’s valuation was driven more by broader concerns about the outlook for the AI industry.
After the U.S. midterm elections, pessimism over tighter regulation of data centers triggered another bloodbath for AI stocks. On November 6, the Nasdaq fell 5.52%, while the Philadelphia Semiconductor Index plunged about 12%.
In this process, the decline in Anthropic’s market capitalization represented the evaporation of its premium for future potential.
Moreover, the alignment with tech giants and reliance on their computing power will also make investors cautious. After all, of the impressive financing figures and soaring revenue on the books, how much comes from organic demand in the real market, and how much is the result of a circular game in which tech giants “invest money with one hand and sell computing power with the other”?
Against a backdrop of total committed spending of up to $90.09 billion under purchase commitments and lease agreements, Jensen Huang abruptly resigned after 33 years at the helm of Nvidia. Combined with the new signals sent after CFO Colette Kress succeeded him as CEO, this removed one of the most critical load-bearing pillars from the AI computing-power financing game.
On Nvidia’s earnings call, Kress publicly laid out the new rules: Nvidia will scrutinize companies receiving financing support with extreme rigor going forward. Data-center guarantees will take effect only after facilities have actually been completed and are ready for operation, with no guarantees provided during construction. If construction of a data center is delayed, Nvidia’s guarantee will cease to exist.
Over the past two years, Nvidia has not merely sold the picks and shovels; it has also served as the “super-guarantor” for the entire buildout of AI infrastructure. As of the end of October this year, Nvidia’s exposure to guarantees for external AI infrastructure financing had snowballed to $11.8 billion.
Previously, the market was willing to assign high valuations to large-model giants such as Anthropic that were deeply tied to computing power, largely because it assumed Nvidia had their backs and would shoulder the burden if anything went wrong. Now the leader has suddenly changed, and the newcomer’s first move was to withdraw the blank check promising unlimited support. It is difficult for the market not to worry.
With that credit backing no longer effective, everyone must reassess whether Anthropic’s future growth expectations can be realized—and, if so, to what extent.
The same is true in the Chinese market.
Over the past two years, many Chinese large-model companies have secured high valuations by tying themselves to major tech companies’ cloud resources and obtaining strategic investment. But Anthropic’s experience has already offered a preview of the outcome. Once they enter the public market, investors will sooner or later separate the boost from major tech companies’ subsidies from actual commercialization performance and evaluate each on its own.
Anthropic’s listing has sent a clear signal to Chinese large-model companies: The era when a story of rapid growth in a single dimension could win over the entire secondary market is already over.
Watching Anthropic fall from the lofty heights of a $2 trillion valuation back to the reality of $1.5 trillion inevitably raises concerns: With Chinese large-model companies likewise relying on revenue growth to drive up their valuations, what fate awaits them next?
Valuation Deflation, Demand Frenzy
How Should Chinese Large-Model Companies Be Priced?
To understand how Chinese large-model companies should be valued, we must begin with the changes at Anthropic, because what changed was not Anthropic itself but the yardstick used by capital markets to price it.
In fact, compared with Wall Street only now beginning to “squeeze the froth” out of Anthropic, the market capitalizations of Chinese large-model companies had already moved to more rational levels.
Take the currently listed Zhipu AI and MiniMax as examples. Even before Anthropic rang the opening bell, the two companies had already demonstrated the process of “valuation deflation” for the entire industry.
When Zhipu AI and MiniMax listed on the Hong Kong Stock Exchange in January this year, they were the toast of the market, with Zhipu AI even setting a record by reaching a market capitalization of HK$1 trillion. Why were they so outrageously expensive at the time? It came down to two things: scarce supply and a narrative of soaring ARR.
When MiniMax went public in Hong Kong, only about 17.07 million of its shares were in public circulation, accounting for less than 6% of its total share capital. Yet the offering drew frenzied subscriptions from 420,000 investors and was oversubscribed by as much as 1,837.17 times. Zhipu AI likewise recorded an oversubscription rate of 1,159.46 times.
But just a few months later, their current valuations tell a different story: Both companies’ market capitalizations have fallen by more than 80% from their peaks. Business revenue is still soaring, but as the expiry of lock-up periods approaches, the scarcity premium on their publicly traded shares has vanished in an instant.
Zhipu AI’s Market Capitalization
More concerning are the large-model companies still at the table but not yet listed—the chill in the secondary market is inevitably flowing back into the primary market.
OpenAI, the overseas industry leader, has also recently been moving at full speed to advance a new funding round in order to replenish its war chest before Anthropic’s listing. OpenAI had originally aimed to raise $30 billion at a valuation of $1.4 trillion, but investors clearly did not have that much appetite.
Ultimately, the valuation for this funding round was cut from $1.4 trillion to $1.25 trillion, wiping out $150 billion—a decline of 10.7%—while the initial amount received also shrank to around $15 billion. Under investors’ scrutiny, OpenAI could not avoid a single bit of the discount it deserved.
If even OpenAI must voluntarily bow its head and accept a discount, then turning our attention back home, how much room do China’s still-unlisted large-model companies have left to raise their valuations?
Once the froth has been squeezed out, the value of China’s large-model companies will no longer depend on grand promises, but must instead be tied to their actual ability to commercialize and generate revenue.
Over the past two years, China’s AI sector has been deeply anxious about the commercialization of large models: revenue from B2B MaaS has been limited, while the consumer market has yet to produce a genuine blockbuster product.
By 2026, with the boom in coding and demand for agents taking hold, token prices began to rise.
According to the Large-Model Token Service Performance Monitoring Report (2026) published by the China Academy of Information and Communications Technology, the normalized average domestic price was RMB 6.8 in the second half of 2025, compared with RMB 13.46 since the beginning of this year, an increase of nearly 100%. Overseas prices nevertheless remain higher than domestic ones, with representative models released since 2026 averaging a normalized price of approximately RMB 117.8.
Why are people still rushing to use them despite the price increases? Because once large models genuinely penetrate complex production processes such as software development and enterprise-grade agents, customers are no longer looking for cheap chat toys, but productivity tools that can directly perform work in place of people. As long as a model is smart enough and can reliably complete workflows, companies are willing to pay real money for it.
During Nvidia’s earnings call, the interim CEO disclosed a new set of figures: rental prices for B300 chips based on the new-generation Blackwell architecture have surged by more than 60% from when they were first launched. Even older cards based on the previous-generation Hopper architecture are commanding multiyear renewal prices as high as twice their original price. Whether hyperscale cloud providers or emerging computing-power cloud companies, all of their computing capacity is running at full load.
Domestic computing capacity also remains in short supply. On November 18, Alibaba CEO Eddie Wu said during an earnings call that as production capacity ramps up, domestically produced computing capacity will continue to account for a larger share of the company’s overall computing capacity and become the main engine supporting cloud revenue growth.
Next, as supply-chain constraints intensify and competition among major companies for accelerator cards reaches fever pitch, computing cards on the market will only become scarcer in 2027. When steadily growing application demand collides with an even tighter supply of computing power, token prices will most likely continue to rise.
Whether it is the free-float game in the Hong Kong stock market or surging valuations in the primary market, whenever valuations run ahead of actual commercial results, the secondary market will sooner or later squeeze out the froth in the most brutal way possible.
The players remaining at the table must use their “pricing power” over tokens to prove that they are irreplaceable. The fact that token prices rose in 2026 even as usage volumes surged has already shown China’s large-model companies a path to survival: whoever can make enterprises remain dependent on its API even after a price increase will have secured genuine pricing power.
What Anthropic has sounded is the death knell for the old pricing era. As the tide recedes, large-model companies that have been swimming naked need to reassess exactly how much they are worth.
“Special Notice: The above content, including any videos, images, or audio, was uploaded and published by users of Dafeng Hao, a self-media platform operated by Phoenix New Media. This platform merely provides information storage services.
Notice: The content above (including the videos, pictures and audios if any) is uploaded and posted by the user of Dafeng Hao, which is a social media platform and merely provides information storage space services.”
AI Trends