Anthropic’s path to an IPO accelerated after Claude Opus 4.5 pushed its Claude Code coding agent into end-to-end software building—paired with fewer usage caps and a more developer-friendly desktop interface. The result is a fast climb in enterprise revenue ru
On Monday, Anthropic submitted a confidential draft registration statement—an S-1—to the U.S. Securities and Exchange Commission, stepping onto the IPO track.
The filing lands after a shift that started late November, when Anthropic released Claude Opus 4.5. At the time. the company was playing second fiddle to OpenAI. with a lower valuation and far less attention in the early generative AI boom. Opus 4.5 changed that momentum by giving Anthropic’s Claude Code coding agent a new “brain” and pushing it toward being described as an “AI killer app.”.
Developers had been using Claude Code to build software for much of 2025. but it had shown more promise than clear. game-changing results. Opus 4.5 is what moved it from “promising” to something engineers could actually run with: it enabled end-to-end app or feature building from plain-language planning and user guidance prompts.
It also made Claude Code’s work longer-running and more structured. The tool improved its ability to discuss a project with an engineer-user, present a plan, incorporate feedback, and then carry out a focused set of multistep tasks to complete a software build.
For teams that live inside deadlines, Anthropic’s product tweaks mattered as much as the underlying intelligence. The company imposed fewer usage caps, a point that can be decisive for software engineers deploying multiple agents to build different parts of a project.
Then there was the interface. Instead of forcing engineers to use Claude Code through their machine’s command line interface. Anthropic gave them a Code tab in the Claude desktop app with its own integrated terminal. Under that tab. it bundled resources developers typically rely on separately—file editor access for viewing and editing code. code-change review windows. parallel coding sessions for different tasks. and other tools. In effect, Anthropic changed Claude Code from a terminal/chat tool into something closer to a full desktop coding environment.
Pieter Levels. a Dutch entrepreneur and self-taught developer with a large software following. publicly celebrated the shift in functionality and time savings. He tweeted, “Claude Code is excellent at fixing all those small bugs from old projects . . . ” adding that “Before I’d not have the time to fix these kinds of projects . . . but now it takes me an hour to do this and it works again!”.
That kind of real-world usefulness fed another expansion: engineers didn’t have to leave their primary workflow to use Claude. The unified interface in the Claude desktop app invited software engineers to do nonsoftware development work in the same place. With easy access to Claude chat and CoWork. engineers could use Claude for research. accessing company data. composing emails. and styling presentations.
The setup also served workers who code only occasionally—people who might build a prototype app to pitch to colleagues, but still spend their regular workday relying on Claude chat and CoWork’s data connections, skills, and workflows.
Anthropic’s accelerating enterprise momentum can be traced back to Claude Opus 4.5. At the end of 2025, the company reported $9 billion in annualized revenue run rate. By February, it reported $14 billion in ARR, and by April it reported $30 billion. A May report pegged Anthropic’s ARR at $47 billion.
OpenAI’s reported performance provides the comparison: at the end of March, OpenAI said it was generating about $2 billion per month in revenue—roughly an ARR of $24 billion.
The company’s enterprise pull appears to be widening too. Anthropic reportedly said that more than 1. 000 companies are paying over $1 million per year for Claude. and that this number has more than doubled since February 2026. Anthropic has also announced major deployments with PwC, Allianz, Snowflake, Accenture, Deloitte, and IBM.
Financial expectations are pushing toward profitability, at least on paper. Anthropic expects to make an operating profit of $559 million during the quarter ending in June, on revenue of $10.9 billion, according to The Wall Street Journal.
But adoption data also suggests the addressable market still isn’t fully captured. While Anthropic does not release raw user numbers for Claude Code, a survey can point to runway. An early-April JetBrains survey of 10,000 software developers globally found 29% used GitHub Copilot, while only 18% used Claude Code, tied with Cursor.
Now, the timing of Anthropic’s move into public markets is starting to look like a coordinated rush. Anthropic’s draft registration statement could make it the second of three major AI-related IPOs in 2026. SpaceX has already filed, and OpenAI is expected to file later in the year.
Both Anthropic and OpenAI have a chance to IPO at trillion-dollar-plus valuations. placing them among the largest tech IPOs in history. Anthropic’s fundraising backstops that scale: the company raised $65 billion in new financing at a valuation of $965 billion. including the new money. just last week.
Still, the story that will matter most to investors isn’t whether Claude Code can be adopted in enterprise settings. Experts aren’t worried about that. The more serious threat, for both Anthropic and its customers, is what it costs to run.
Claude Code’s appeal depends on deep understanding of code bases and complex coding problems. That requires the agent to reason across large context windows, which in turn means consuming a lot of tokens—the chunks of text, data, and code it processes.
Corporate users are already feeling the strain. Uber is among the companies running up big bills as engineers max out on tokens. Anthropic, facing the same reality at scale, is trying to find additional computing resources to process the tokens generated by Claude Code users.
The company is even paying Elon Musk $1.25 billion a month for use of the xAI/SpaceX Colossus 1 data center in Memphis.
By the time Anthropic’s prospectus arrives—likely sometime this summer—the market may finally get the answer it’s been waiting for: the real profitability of selling Claude Code, and whether the economics of tokens and compute can keep up with the product’s momentum.
Anthropic Claude Opus 4.5 Claude Code IPO SEC S-1 enterprise AI annualized revenue run rate ARR token costs xAI/SpaceX Colossus 1 Memphis data center PwC Allianz Snowflake Accenture Deloitte IBM Uber
So they’re doing an IPO now? That sounds like hype not product.
I don’t get it, I thought Claude already code’d. Like what makes 4.5 different, did they just remove the “caps” so people burn more credits?
“AI killer app” is wild marketing lol. Also if it can build software end-to-end from prompts, doesn’t that just mean it’ll replace devs faster? Or is that what they’re not saying in the S-1 draft?
Confidential draft registration statement sounds like they’re basically already selling shares, right? Or is it just paperwork to make the valuation go up. Reminds me of when OpenAI got all the spotlight and Anthropic was in the background… now all of a sudden they’re catching up because the desktop interface is “developer friendly”??