OpenAI 拟以 1.4 万亿美元估值融资 300 亿美元，抢在 Anthropic IPO 前

OpenAI is accelerating its planned $30 billion fundraising round ahead of rival Anthropic’s expected initial public offering (IPO) in mid-November, a move that could value the ChatGPT maker at more than $1.4 trillion, according to a Reuters report.

SoftBank Group, Abu Dhabi-based MGX, and other existing OpenAI backers are in talks to participate in the round, according to the report.

OpenAI held meetings with long-only investment funds and sovereign wealth funds last week, while Anthropic hosted its own investor day.

Per the report, OpenAI is pitching investors on revenue growth following the launch of its GPT-6 Astra model. The company has told investors that its annualised revenue has reached $50 billion and could rise to $70 billion by the end of 2026, according to Reuters.

OpenAI was last valued at $852 billion in March 2026, following a $122 billion funding round that included a commitment from Amazon to invest up to $50 billion in the company.

Anthropic, meanwhile, recorded annualised revenue of around $65 billion by July. Its quarterly revenue increased to $16.8 billion in the third quarter of 2026, from $11.6 billion in the second quarter.

It reported an $80.7 billion net loss and disclosed more than $600 billion in future computing infrastructure commitments, according to The Wall Street Journal (WSJ), which reviewed the draft IPO prospectus.

Anthropic’s initial public offering (IPO) could value the company at as much as $2 trillion.

Anthropic’s net loss for the six months ended June 30, 2026, was substantially higher than the $13 billion loss recorded in the same period last year. Much of the latest loss stemmed from accounting adjustments linked to financial instruments used to raise capital, rather than operating expenses alone.

The company’s infrastructure spending also increased sharply as demand for computing power grew. Anthropic spent nearly $9 billion on computing capacity and infrastructure in the first half of 2026, compared with $2.3 billion in the corresponding period last year, according to WSJ.

Its first-half spending also exceeded the $7.3 billion it spent on infrastructure throughout 2025, when it generated $4.6 billion in revenue.

Anthropic’s expansion is tied to its ability to secure enough computing resources to train and operate its AI models.

Anthropic and AI infrastructure company Nscale are reportedly preparing a second $45 billion computing agreement, according to Bloomberg.

The company has committed to approximately $450 billion in cloud computing agreements over the next seven to 10 years, with most of these contracts being non-cancellable.

Under the disclosed agreements, Anthropic plans to spend $111 billion on Google Cloud and $110 billion on Amazon Web Services (AWS), alongside $31 billion on Microsoft services. These commitments apply regardless of how much computing capacity the company ultimately uses.

Anthropic has also agreed to spend up to $86 billion renting NVIDIA chips through AI infrastructure providers Nscale and Lambda, with the agreements extending into 2033 and 2034, respectively.

Anthropic’s infrastructure commitments exceed $600 billion, including more than $160 billion tied to leasing Google chips through an arrangement involving Broadcom, according to WSJ. The company is also in talks for additional computing deals worth up to $150 billion.