Anthropic Raises $81.3B in Record IPO at $1.5 Trillion Valuation

By IT之家Original title: Anthropic 完成史上最大规模 IPO,上市估值达 1.5 万亿美元
88AI Score

Anthropic listed on Nasdaq on November 19, raising $81.3 billion in the largest IPO in history and reaching a $1.5 trillion fully diluted valuation, becoming the first frontier AI lab to go public. Revenue hit $16.2 billion in H1 2026 on explosive growth, but losses reached $80 billion over the same period, with compute the largest expense; its post-listing performance will serve as a key bellwether for the multi-trillion-dollar AI investment boom.

Full Report

IT之家 November 19 news: Late tonight Beijing time (the 19th), AI startup Anthropic officially listed on the U.S. stock market, setting a historic IPO record with $81.3 billion (IT之家 note: approximately 543.147 billion yuan at current exchange rates) in proceeds.

CEO Dario Amodei rang the opening bell at Nasdaq. "Today's listing makes the company's financial foundation more solid and strengthens our determination to do the right thing."

Anthropic's offering price was set at $83 per share (approximately 554.5 yuan at current exchange rates), above the expected range announced last week, giving it a fully diluted market value of $1.5 trillion (approximately 10.02 trillion yuan at current exchange rates).

Bloomberg and Reuters reported that Anthropic's stock price is expected to rise about 20% after official trading begins, reaching around $100 per share (approximately 668.1 yuan at current exchange rates).

On the 18th local time, Anthropic also announced the sale of 980 million shares to underwriters, with the expanded offering size bringing total proceeds to $81.3 billion (approximately 543.147 billion yuan at current exchange rates), breaking SpaceX's record of $75 billion (approximately 501.058 billion yuan at current exchange rates) raised in its June IPO this year.

Founded in 2021, Anthropic has taken only about five years to be poised to join the ranks of the world's most valuable listed companies.

Anthropic is the first listed frontier AI research organization, and its stock performance will serve as an important bellwether for observing the trillions of dollars of AI investment enthusiasm. Wall Street investors and the global tech industry are closely watching the company's post-listing performance.

Anthropic's prospectus, filed at the end of October, shows that the company's revenue has grown explosively every year since its founding, with some years seeing increases exceeding 1000%. In the first half of 2026, Anthropic's revenue reached $16.2 billion (approximately 108.229 billion yuan at current exchange rates), far above the full-year 2025 figure of $4.6 billion (approximately 30.732 billion yuan at current exchange rates).

But Anthropic has also paid a huge financial price. The company lost $80 billion (approximately 534.462 billion yuan at current exchange rates) in the first half of 2026, mainly due to a one-time charge, with cumulative losses widening further to $104 billion (approximately 694.801 billion yuan at current exchange rates) over the first nine months.

As Anthropic continues to expand its computing resource investment, costs are expected to climb substantially. Computing power is mainly provided by semiconductor chips typically deployed in data centers, and is the foundation for running generative AI and Anthropic's largest expense item. In the first nine months of 2026, the company spent about $18.8 billion (approximately 125.599 billion yuan at current exchange rates) on computing power.

Anthropic executives wrote in the prospectus: "Developing frontier AI models requires substantial investment in computing power, capital, and manpower, involving large-scale training that may last weeks or even months, with each training run potentially costing millions of dollars."

Anthropic also disclosed in detail in its offering documents the various risks its AI models may pose to humanity, including being used to develop biological weapons or exacerbating oppression around the world. Previously, employees had worried that some companies were developing advanced AI models irresponsibly, and Amodei subsequently called for slowing the pace of development of the most advanced AI models.