Stripe buys OpenRouter for $7.5 billion (๐ณ)
This is a translation of a Norwegian article I initially published on LinkedIn, August 28 2026
What does it mean for OpenRouter's future, and especially for those of us who use the platform?
Who is Stripe?
Stripe is the fifth-largest payment processor in the US. It's also one of the companies developers respect most, known for reliable APIs of very high quality. It has positioned itself as a provider of neutral infrastructure, considerably lowering the barrier for anyone who needs a payments platform. In 2025, Stripe processed a total of $1.9 trillion in payments (1,900,000,000,000).
Who is OpenRouter?
OpenRouter offers over 400 different AI models from more than 80 different AI providers, to 8 million customers.
It thus serves as an integration point between developers/businesses and the various models/providers. OpenRouter makes you less exposed to providers' uncertainty and volatility (see the withdrawal of Fable), and to price changes from individual providers.
The basis of OpenRouter's trust
And precisely because that is its business model (rather than making/providing the best model), people see it as neutral. Customers trust OpenRouter because it's clear what the company is and what it provides. That neutrality is an important reason it has grown so big so quickly.
How quickly?
In May, it raised $113 million in a funding round at a valuation of $1.3 billion. In other words, a 5-6x increase in value in three months, based on Stripe's winning bid.
What changes with Stripe as owner?
Chinese models account for almost half of US companies' AI traffic, which flows through a gateway Stripe now controls, with real exposure to regulatory requirements and compliance. That will probably hit OpenRouter harder with Stripe as its owner.
But above all, this is about trust.
An example: OpenRouter provides a service called Auto Router. When you choose openrouter/auto as your model, you're automatically routed to the model OpenRouter thinks is best suited to your task. It makes that assessment by analyzing customers' tasks and model choices, using a dataset of 55 trillion (yes, trillions again) tokens per week. Here, "best suited" means "the model most popular for the task class into which it has placed the incoming task."
Using Auto Router instead of choosing a model yourself therefore means trusting OpenRouter to base its choice on what it believes is the best model for you at that moment. In this case, "most popular" is a proxy for "best," but it's still based on what's best for you.
Will this calculation continue to be based on what's best for you?
You don't have to do anything wrong to lose trust. It's enough for uncertainty about motives, neutrality and so on to arise for trust to weaken. A good example is the law on conflicts of interest. Lawmakers have explicitly established that the possibility someone might doubt a judge's or public official's impartiality is enough to trigger the rules on disqualification. The possibility alone is enough to sow doubt.
Stripe already has deep commercial ties to many of the AI providers. If Stripe processes payments for Anthropic, OpenAI and Google, and also controls the distribution layer that determines how much AI traffic each receives, it's easy to imagine conflicts of interest arising. Can we trust that commercial considerations won't find their way into the algorithms governing which model your task is sent to, or will doubt arise?
For a company whose central foundation is largely trust, that doubt can have serious consequences.