New York, June 10, 2026, 15:41 (EDT)
- Apollo and Blackstone are putting up money for Anthropic’s $35 billion push to boost AI computing, with Broadcom designing custom chips for the work. The price tag shows how costly the IPO pitch for the Claude developer is growing.
- Anthropic filed a confidential draft S-1 for an IPO on June 1, just after raising $65 billion at a $965 billion post-money valuation.
- OpenAI’s confidential S-1 filing has pushed the Anthropic IPO into a wider contest to grab public-market AI cash.
Anthropic’s possible IPO isn’t just about Claude demand anymore. The last 48 hours shifted the story—now it’s about new financing. Apollo and Blackstone are putting up money for a $35 billion AI-compute expansion. Anthropic will use Broadcom chips. The move comes as investors weigh if any private AI firm near a $1 trillion valuation can turn big user numbers into lasting profits.
Anthropic, maker of Claude, is moving ahead with IPO plans. The company said June 1 it filed a confidential draft S-1 registration statement with the SEC for a planned public listing. It hasn’t fixed the number of shares or a price range yet. A confidential S-1 lets the SEC review filings before investors get the full prospectus.
Apollo and Blackstone are putting up funds to boost Anthropic’s power supply as AI growth shifts demand from app downloads to actual power plants. The two firms are backing an initial one-gigawatt addition at Fluidstack-run sites by mid-2026, Reuters said Tuesday. That’s about the same as what it takes to run 750,000 homes. The broader project looks to top 20 gigawatts of computing capacity for major AI labs, like OpenAI, through 2028.
Scale works both for and against Anthropic. It helps answer investor concerns about capacity problems for model companies. But it also raises the stakes for a future public filing—shareholders will want to know how much of Anthropic’s revenue is going to chips, power, leases, and long-term cloud deals.
Anthropic is pitching growth numbers that stand out even in the AI world. The company said in late May it pulled in $65 billion through a Series H round led by Altimeter Capital, Dragoneer, Greenoaks and Sequoia Capital. That deal put its valuation at $965 billion with the new cash added in. Anthropic reported run-rate revenue over $47 billion earlier in May — that’s not audited annual numbers but revenue paced over 12 months. “The funding will help us serve the historic demand we are experiencing,” CFO Krishna Rao said. Anthropic
Anthropic pushed out Claude Fable 5 on June 9, describing it as its most capable generally available model so far. Some cybersecurity, biology, chemistry and model-distillation queries will be routed to Claude Opus 4.8 instead under new safeguards, the company said. The move aims to show IPO watchers that Anthropic can keep advancing its models while expanding paid access.
Fable 5 brought up a new enterprise issue. Reuters said Wednesday that Microsoft is restricting staff access to Claude Fable 5 while its lawyers look at Anthropic’s data-retention terms. The rules for Mythos-class models say prompts and outputs stick around 30 days for trust and safety, or up to two years if a classifier flags them for use violations.
Anthropic is taking different approaches to tackle the compute issue. In April, the company announced a fresh deal with Google and Broadcom to get multiple gigawatts of next-gen TPU capacity beginning in 2027; TPUs are custom AI chips from Google. Also in April, Anthropic said it planned to spend over $100 billion on AWS tech across 10 years in a separate Amazon agreement. That deal promises as much as five gigawatts of capacity.
That’s why public markets are watching even before Anthropic gets a ticker. If it lists, it would give investors a rare shot at backing a frontier AI model developer, instead of buying shares in suppliers, cloud firms, or big investors in the sector. BCA Research chief strategist Noah Weisberger told Business Insider that more AI IPOs could “reduce scarcity value” in current AI names as money moves to stocks with direct exposure. Business Insider
OpenAI is turning up the heat on the rotation. The ChatGPT developer confirmed it filed a confidential S-1, and Reuters said OpenAI is eyeing a valuation reaching $1 trillion if it moves ahead with a public listing. Reuters also cited Michael Ashley Schulman from Cerity Partners, who said OpenAI is keeping its options on the table after Anthropic filed first.
There’s a risk the IPO window could shut before Anthropic gets to market, or numbers might not hold up once public investors take a closer look. Anthropic hasn’t given an offering size, price range, or a date. The company said the IPO timing will be driven by market conditions. The SEC requires companies using nonpublic review to release the registration statement and draft S-1 at least 15 days before a roadshow, or before the planned effective date if there isn’t one.
AI IPOs now face rising political risk. Reuters said Wednesday that President Donald Trump expects major AI companies to talk about “giving back” to the public, possibly through government stakes. According to Reuters, Anthropic, OpenAI, Google and Meta didn’t reply to comment requests. Reuters
The next big event is the public S-1. Investors will find out if Anthropic’s $47 billion run-rate and recent model launches are enough for the company’s huge compute costs — or if the IPO turns into Wall Street’s first close look at just how pricey the frontier AI push is getting.
