From OpenSea to OpenRouter: Is Alex Atallah’s “high-position exit” script being replayed?

Author: Nancy, PANews

More than four years ago, Alex Atallah exited before the peak of the NFT bubble; now, he is again stepping into the spotlight amid the AI boom, preparing to sell his AI model aggregation platform, OpenRouter, for a high price.

On July 23, according to The Wall Street Journal, payment giant Stripe is in talks to acquire OpenRouter, with the deal valuation possibly nearing $10 billion. If the deal ultimately goes through, it would mark another success for Alex Atallah in building a company at the billion-dollar scale after OpenSea.

Stripe to acquire OpenRouter, valuation may hit $10 billion

The rumor that OpenRouter is looking to “sell itself” has been simmering for quite some time.

Last week, multiple overseas outlets including The Information and Jawl reported that OpenRouter has received acquisition intentions from several large tech companies and is discussing the sale; the potential deal size could be as large as several billion dollars.

And according to the latest update from The Wall Street Journal, Stripe is considering acquiring the world’s largest AI model aggregation platform. Insiders said that the negotiations between the two sides may be announced soon, but there are still uncertainties in the deal right now—there is no exclusion that talks could fall apart or that other potential buyers might join the bidding.

Interestingly, Alex Atallah previously described OpenRouter as “Stripe in the AI space.” In his view, just as Stripe helps customers handle various payments through a unified entry point, OpenRouter aims to become a unified entry point for enterprises to access different AI models, reducing the cost for companies to switch among multiple AI vendors, while also avoiding being “locked in” by a single model supplier.

In fact, OpenRouter has already established cooperation with Stripe long ago. This potential transaction is also seen as an important move by the payments-infrastructure giant to further extend into the AI infrastructure space. It’s worth noting that this could be another major acquisition action by Stripe in the near term—around the same time, market rumors said Stripe plans to acquire PayPal.

At present, neither side has disclosed the specific deal amount. However, according to insiders, if the transaction is ultimately completed, the valuation at which OpenRouter would be sold may approach $10 billion.

That figure would far exceed OpenRouter’s valuation when it previously raised funding. After just more than three years since its founding, the AI infrastructure company has achieved rapid growth, driven by the big model wave.

Public data shows that OpenRouter has completed three rounds of funding, cumulatively raising more than $150 million. In June 2025, OpenRouter announced the completion of a $40 million seed and Series A round, with a post-investment valuation of about $547 million at the time; on March 26, 2026, OpenRouter announced the completion of a $113 million Series B round, with a post-investment valuation of about $1.3 billion.

If this deal eventually closes, it would mean OpenRouter’s valuation could rise nearly 10x within just a few months, and it would join the club of $2M-plus “unicorns.”

Behind OpenRouter’s rapidly rising price tag is not only the demand surge brought by the fast expansion of large models, but also the fact that the AI infrastructure track is receiving more attention from capital markets.

Replaying the OpenSea exit playbook? Why did OpenRouter choose to “sell itself”

This is not the first time Alex Atallah has built a $10 billion-level company.

Before that, the serial entrepreneur co-founded OpenSea and brought NFTs from a niche circle into the mainstream. During the NFT craze, OpenSea grew from a little-known platform into the world’s largest NFT marketplace, with a valuation once exceeding $13 billion, and the two founders’ net worth also once reaching about $2.2 billion.

However, before the NFT market fully cooled down in 2022, Alex Atallah chose to leave OpenSea. After the industry bubble burst, OpenSea’s valuation fell sharply, and the former NFT leader gradually lost the shine it had during its peak. And Alex Atallah’s early exit was also once viewed by the market as an important signal at a major turning point.

Afterward, Alex Atallah turned his attention to AI infrastructure, and the OpenRouter he founded became the biggest hub for the AI era.

Currently, OpenRouter has integrated more than 400 AI models, has about 10 million users, and processes over 200 trillion tokens per month. Since the start of this year, the number of tokens processed through its API has grown by about 10x.

But it’s worth noting that despite its rapid business growth, OpenRouter has not chosen to go public (IPO)—instead, it is heading toward a potential sale. The reason is that it is a business with a huge scale, but limited profit margins.

At present, OpenRouter mainly makes money by charging platform service fees when developers call AI models, taking a cut of about 5%-5.5%. Although its annualized AI inference spending scale has already reached several hundred million dollars, as of April 2026, the company’s annualized revenue is about $50 million.

In other words, OpenRouter connects a massive AI demand market, but it does not fully control the value-chain upstream. As model capabilities gradually become standardized, the platform is more vulnerable to factors such as the rise of open-source models, the cloud vendors’ ecosystem lock-in, and model suppliers directly lowering prices—so profit margins may face sustained pressure. OpenRouter may find it hard to tell an “exponential growth with ultra-high profits” story to the capital markets.

Meanwhile, competition in the AI model aggregation track is also intensifying. For example, Meta’s in-house AI incubator overseas is developing a scheduling service to benchmark OpenRouter, aiming to reduce code development compute costs; in China as well, large-model aggregation platforms have emerged, including Cheetah Mobile EasyRouter and NetEase Youdao ThinkFlow.

Therefore, the market’s high valuation for OpenRouter is more about pricing in its future imagination space, rather than its profitability capacity at the current stage.

However, for potential buyers, OpenRouter’s truly attractive asset may not be its current revenue scale, but the real AI usage data it has accumulated over the long term.

By connecting hundreds of models and tens of millions of users, OpenRouter has accumulated a large amount of call data from real production environments—covering differences in how various models perform in actual tasks, developers’ preferences, price sensitivity, and substitution relationships between open-source and closed-source models. Compared with lab test data, this real-world AI usage data is closer to market demand and is also harder to replicate quickly through short-term investment. This may be an important reason why large tech companies are willing to pay a substantial premium for OpenRouter.

From NFTs to AI, Alex Atallah has twice caught the winds of the era. If OpenRouter is ultimately sold with a $10 billion valuation, does it mean a re-pricing of the value of AI infrastructure—or is it yet another signal of the peak of a cycle? The answer may still require time to be verified.