What if there's only One Major AI Winner?
Also: Claude Cowork Plugins: A Beginner's Guide.
Good Morning,
As we get ready for the major artificial intelligence IPOs of 2026, there’s a stark realization dawning on many of us. Given that two of the companies going public have profitability and revenue issues, there’s no way around the conclusion - there will likely be one major winner in all of AI.
Remember, Anthropic is on track to generate $10.9 billion in revenue during the second quarter and it just generated $4.8 billion in revenue during the first quarter. How much will it even be making by the time it goes public in October or November, 2026? That’s just around 4.5 months away. The numbers don’t lie.
Anthropic is now likely generating at least 35% more revenue than OpenAI, reversing the companies’ positions from the end of last year. - The Information
Kindly share this article if you think this is important AI News that someone you know might enjoy reading about.
There is no duel with SpaceX or OpenAI ahead, Anthropic has already won. SpaceX has become the Neo Cloud of Anthropic and OpenAI will have to contend with a Meta (formerly Facebook) that now is pushing subscriptions on its AI app. OpenAI has inferior product and AI app execution compared to ByteDance, among others. This isn’t a model or compute story, this is a product execution and business strategy story.
I Predict a $1.4 Trillion IPO for Anthropic
At Anthropic’s current pace, by the time it goes public that could be for $1.4 Trillion of a market cap. The average business is spending 13x more on AI tokens (in May, 2026) than in January 2025, according to Ramp data just six months later. Meta’s only way to pay for its outrageous AI capex, is to start a Cloud computing provider. “It’s definitely on the table,” Zuckerberg said at Meta″s annual shareholders meeting. I don’t see a future where Anthropic and OpenAI don’t also own and operate their own Cloud computing businesses as well by 2030. It’s necessary to optimize revenue and lock down partnerships.
Meanwhile ByteDance says it will triple its AI capex in 2026 to an astounding $70 Billion. What one quickly realizes is the following: There won’t be many winners in AI in such an environment, there will be just a few big winner. There will also be one major winner taking the bulk of global revenue in the years ahead. As impressive as ByteDance and Alibaba have been and are in AI , they don’t have the Enterprise AI chops, models and credibility of Anthropic. Google’s I/O conference was very disappointing, while competitive, Alphabet is unlikely to be the best AI company. It’s an Advertising company primarily with a few other aspects. Its innovation is primarily defensive. It’s an incumbent, and no matter how talented the people there are, many of them will move on to greater things.
There’s no one left to compete with, Anthropic by the end of 2026 is in a tier of its own making.
Paid subscriptions is not going to pay for Meta’s morbidly increasing AI related Capex. In April, 2026 Meta raised its 2026 guidance for AI-related capital expenditures to between $125 billion and $145 billion, up from a prior range of $115 billion to $135 billion. OpenAI (or Amazon for that matter) can spend however much it wants, it’s not going to catch Anthropic that is a pure-play B2B Enterprise AI firm. We are seeing this with the ARR trends so far in 2026.
Anthropic will take just five years to become profitable, 2021 to 2026.
SpaceX will need to merge with Tesla to ever be profitable and OpenAI may never reach that point. I don’t believe OpenAI can ever catch up and I’m somewhat bearish on its ability to compete with the remaining companies.
According to Polymarket: there's now an 93% chance Anthropic has a higher valuation than OpenAI this year. In private markets Anthropic is already worth much more than OpenAI. Its ARR growth is obviously more special and its product execution much more aligned with what Enterprise customers actually need.
This is an AI Supremacy Story of 2026
Anthropic isn't just growing revenue faster than OpenAI, Google or others—it is compounding at a rate that is fundamentally altering the long-term timeline of how much revenue it will make compared to everyone else.
If historically Peter Thiel was one of the very first and most critical early investors in DeepMind (before it was acquired by Google and became Google DeepMind).
But Demis Hassabis was one of the first and most critical early investors in Anthropic. That is, the leader of Google’s AI has incentives for Google not to win, but rather, for Anthropic to win.
Anthropic is part of an AI oligarchy where both Alphabet and Amazon are major investors. Anthropic can determine the price of its tokens because it has the best AI coding and agentic models. Claude is expensive for a reason. Mythos is too good for even public release. Trends like tokenmaxing simply make Anthropic’s revenue surge of 2026 look more impressive. This is not by accident, or coincidence.
While SpaceX and OpenAI have many structural and execution risk issues ahead, with a morbid amount of competition. Anthropic is the manifest destiny of the highest AI researcher and engineering talent concentration in the world, with among the lowest churn. Anthropic’s ARR difference over OpenAI becomes considerable the further we we get after the IPO of 2026. Anthropic’s margins are so much better than OpenAI, it’s not even remotely competitive.
Anthropic’s Profitability is a Cut Above 🔬👀🌊
 | A guest post by
|