Anthropic Eyes $900B Valuation in $50B Round as Revenue Hits $45B

According to the Financial Times, Anthropic is considering a funding round that would value the Claude maker at around $900 billion pre-money and raise up to $50 billion. This would put the company ahead of OpenAI's $852 billion post-money valuation from March and set it up for what many investors call a "blockbuster IPO" that could happen as soon as the end of 2026.

The number that backs up the valuation: Anthropic's annualised income is about to pass $45 billion, which is five times what it was at the end of last year ($9 billion). If that trend continues, it changes everything we thought we knew about where Anthropic fits in the competitive AI world.

The Revenue Surge That's Driving Investor Demand

The main fact of this financial story is how quickly Anthropic's income is growing. From $9 billion a year at the end of 2024 to $45 billion a year in just over five months is not minor growth; it is a big shift in how businesses use AI.

The goods that are making it happen can be seen. Anthropic's AI tools for software developers, Claude Code, and Cowork, created for business users who aren't as tech-savvy, have become the best in their respective fields. This has helped Anthropic catch up to OpenAI in the business client market.

"Any amount of money is welcome at Anthropic." A business owner told the FT, "It's almost time for them to pop their heads up and say 'we're ready.'"

Dragoneer Investment Group, General Catalyst, and Lightspeed Venture Partners are some of the companies that are interested. Existing backers are asking to be allocated in the new round, even though CFO Krishna Rao hasn't officially started talks about terms yet.

This shows how demand-driven the process is. There is no agreement on terms yet, and there is no guarantee that the deal will go through on time. The round is due to end in two months.

The Compute Problem — and How Anthropic Solved It

It's not mainly an assessment exercise that's being done to raise money. Anthropic can no longer put off this plan to use cash to expand its computing infrastructure. Recently, the company hasn't been able to serve customers as well as it would like to because of limited supply. Also, the upcoming wider release of Mythos, Anthropic's powerful new AI model that is currently only available to a small group of partners, will require a lot more computing power than the company has right now.

Anthropic has taken a strong stance against this. It has made deals with SpaceXAI, Google, Broadcom, and AWS in the last two months to supply computers. These deals have added hundreds of billions of dollars to its long-term costs in return for guaranteed capacity. The massive $100 billion, ten-year deal with AWS that was just announced is the most important piece of that technology stack. These deals lock in the supply of computers, which buyers see as a way to stay ahead of the competition.

One backer in the company said, "Anthropic has fixed the biggest bottleneck and possible source of weakness, which is compute." It was on purpose that CFO Rao didn't consider higher-value offers until the compute deals and a new private equity partnership were finished. This is an example of a disciplined capital strategy, not a sense of urgency.

For background, OpenAI President Greg Brockman said in court that the company will spend $50 billion on computers just this year. Anthropic is building up to a similar infrastructure position, but it will cost a lot less per dollar if the deals with AWS and Google produce the contracted capacity on time.

The OpenAI Comparison: Valuation Crossover in Real Time

The most interesting thing about the funding round is how it affects competition. In March, OpenAI raised a record $122 billion in funding, giving the company a value of $852 billion after the money was added.

At the time, this was the biggest private funding round ever. With a pre-money goal of $900 billion, Anthropic would be the more valuable company when the deal closes, which would not have been possible twelve months ago.

The difference in value shows how the market thinks about which AI lab is making more progress with businesses faster. Business clients care more about reliability, safety positioning, and developer tooling than consumer brand knowledge.

Anthropic's strength in these areas has proven to be more commercially durable than expected. Businesses using AI agents in work settings are the ones who use Claude, which is what makes the $45 billion a year.

IPO Positioning and Investor Strategy

The size of the round and Anthropic's stated goal to go public are influencing the choice of investors. Several sources say that the company will likely prefer investors who have experience in both the public and private markets. This is done on purpose to build up the pool of big investors that will be needed before the IPO roadshow.

Investors who want to get in now are making a pre-IPO bet at a price that is still below what the company is likely to be worth on the public market if it keeps making the same amount of money. The risk in that calculation is clear: Anthropic is being valued at about 20 times its annualised revenue, which means it needs to keep growing at a very fast rate to support the valuation and doesn't leave much room for any slowdown in the adoption of AI in businesses.

Since OpenAI's March raise, Anthropic's $900 billion round has been the most important sign of value in the AI private market.

The growth in sales from $9 billion to $45 billion a year in just five months is what sets this apart from just speculation about AI. It's clear that this company is very good at managing its capital cycle, as shown by its compute infrastructure strategy, product market fit in workplace AI, and IPO preparation sequencing.

This is especially true for a company that is growing so quickly. The risk is carrying out large-scale tasks like deploying Mythos, delivering compute supplies, and keeping up income growth across the whole business at the same time. If Anthropic succeeds, the $900 billion value before the money was raised looks low now. As the current levels rise, if even one piece stops working, the triple compression could be very bad.