What Stripe’s OpenRouter Deal Means for Embedded Finance

On August 19, Stripe agreed to acquire OpenRouter for a reported $7.5 billion. The deal shows embedded finance is moving beyond payments and into the decision flows that generate them.

OpenRouter is a three-year-old company whose tools route AI tasks to particular models and determine how that usage is billed. As of May, it said it served 8 million developers. By acquiring it, Stripe moves further up the AI procurement chain, from enabling a payment to helping determine how a task will be matched with an AI model and how that usage will be charged.

The value comes from sitting inside the workflow that leads to a payment rather than merely enabling the transaction. With OpenRouter, Stripe gains more than fee revenue: It gets insight into how companies use AI, how AI traffic is routed, which models they choose and how much they spend. In time, that approach could extend beyond AI models to other types of technology and purchasing decisions.

From payments to workflow

Consider what Stripe offered when it was founded in 2010. The company made it easy for online businesses to accept payments without having to put together various services from banks and other payment intermediaries.

OpenRouter anchors Stripe inside businesses' AI-purchasing processes. It's enabling AI-based tasks by routing prompts to models and measuring the resulting usage for billing. It also builds on the company's January acquisition of Metronome, a company whose tools help calculate charges based on product usage.

"Stripe is building the economic infrastructure for AI, and together with OpenRouter we'll help businesses maximize profitability by routing their requests intelligently and spending their tokens efficiently," Stripe co-founder and CEO Patrick Collison said in a blog post.

So Stripe is treating model selection and AI spending as extensions of the financial infrastructure it already provides.

By embedding its infrastructure into as many touchpoints as possible, Stripe extracts value along the way. It earns fees from processing payments. OpenRouter generates revenue from AI usage routed through its platform.

"OpenRouter handles much of the messy stuff: routing requests between providers, automatically falling back when one goes down, and optimizing for things such as price, latency and model quality. It generally passes through providers' inference prices without a markup, instead making money through a 5.5% fee on credits purchased through the platform," wrote Paul Sawers in The New Stack.

The strategic value, however, comes from the data. OpenRouter says it doesn't store prompts or responses unless users opt in, but it retains metadata, including token counts and latency, for each request. More data means improved routing and intelligence on how models perform under various circumstances. That could give Stripe a broad view of which models businesses use and how heavily they use them, knowledge could help it identify other services its customers may need. Since Stripe is already part of their operations, it would be well placed to sell those additional services.

Stripe isn't the only tech company aiming to monetize AI routing. In recent months, Ramp and Cursor also rolled out tools that route AI requests. Meta is reportedly developing a similar offering. It all reflects growing interest in influencing how businesses choose and use models.

The takeaway

Fintech infrastructure providers have a growing opportunity to move upstream, from executing payments to becoming part of how spending decisions are made. They'll increasingly compete to own the workflow steps that determine where money goes. At Great North Ventures, we see opportunities in infrastructure providers whose tools sit at the intersection of financial services and the processes that lead to a payment.