Stripe is making a major move into AI infrastructure.
The fintech giant said Wednesday that it plans to acquire OpenRouter, a startup that helps developers access and route requests across a wide range of artificial intelligence models.
The companies did not disclose financial terms. However, The New York Times, citing a person familiar with the deal, reported that the transaction could be worth approximately $7.5 billion, with about $1.5 billion earmarked for OpenRouter’s founders.
The reported valuation would represent a dramatic increase from OpenRouter’s most recent funding round. The startup raised $113 million less than three months ago at a valuation of roughly $1.3 billion.
Stripe declined to comment on the reported figures.
Why Stripe Wants OpenRouter
OpenRouter has gained traction among developers who want access to multiple AI models without being locked into a single provider.
The platform is particularly relevant to users of open-weight models, including systems developed by Chinese AI companies such as DeepSeek and Z.ai.
These models have attracted developers in part because they can offer lower costs than some proprietary alternatives from companies such as OpenAI and Anthropic.
That puts OpenRouter at an important point in the AI supply chain: helping developers decide which model should handle a request and how much that request should cost.
You might want to read this: Palo Alto Networks CEO Warns AI Could Trigger $1 Trillion Cybersecurity Overhaul
The AI Cost Problem
As the number of AI models grows, businesses face a new challenge.
Models are constantly being launched, upgraded and repriced. Choosing the right system for a particular task can therefore become a balancing act between performance and cost.
Stripe said it has already been working with businesses to improve how they manage AI usage and optimise the cost of tokens—the units commonly used to measure the amount of text processed by AI models.
OpenRouter could give Stripe a deeper role in that process.
Instead of simply helping companies move money around the AI economy, Stripe would gain technology that helps determine where AI workloads are sent and how efficiently they are processed.
Stripe’s AI Ambitions
The acquisition would mark another major expansion for Stripe, which has grown into one of the world’s most valuable private technology companies.
The company was valued at nearly $160 billion earlier this year, driven primarily by its payments infrastructure.
Stripe has also expanded beyond traditional payments. In 2025, it acquired stablecoin infrastructure company Bridge for $1.1 billion, strengthening its position in the cryptocurrency and digital-finance market.
OpenRouter represents another step into a rapidly expanding technology sector.
Building Infrastructure for AI
Stripe CEO Patrick Collison said the company’s broader ambition is to build the financial infrastructure supporting the AI economy.
The OpenRouter acquisition fits that strategy by combining payments and business infrastructure with technology designed to make AI spending more efficient.
For Stripe, the opportunity is potentially significant.
As companies deploy more AI applications, they will need not only models but also systems for routing requests, controlling costs and selecting the right model for each job.
OpenRouter’s Bigger Vision
OpenRouter says its goal extends beyond simply making AI cheaper.
The startup has argued for an ecosystem where multiple models can compete, giving developers access to different systems rather than allowing one model to become dominant simply because it is the default choice.
That philosophy aligns with OpenRouter’s focus on giving developers access to a broad selection of models and providers.
The Stripe deal could therefore bring together two complementary ambitions: Stripe wants to become infrastructure for the AI economy, while OpenRouter wants to become part of the infrastructure through which that economy actually uses AI.
If the reported $7.5 billion price is accurate, Stripe is betting heavily that managing AI usage will become almost as important as building the models themselves.