Stripe is making a major push into the artificial intelligence economy with plans to acquire OpenRouter, a fast-growing platform that gives developers access to a wide range of AI models, including popular open-weight systems from Chinese and U.S. developers.

The fintech company announced the deal on Wednesday without disclosing financial terms. The New York Times, citing a person familiar with the matter, reported that Stripe is paying about $7.5 billion, including roughly $1.5 billion earmarked for OpenRouter’s founders.

The reported price marks a dramatic increase from OpenRouter’s most recent private-market valuation. The startup raised $113 million less than three months ago at a valuation of about $1.3 billion, meaning the reported acquisition price would value the company at nearly six times that level in a matter of weeks.

Stripe declined to comment on the reported valuation.

The acquisition gives Stripe exposure to a segment of the AI market that sits between model developers and businesses that use AI. OpenRouter acts as a routing layer, allowing developers to access and compare multiple AI models rather than building their applications around a single provider. That position is becoming more important as companies face a rapidly expanding selection of AI models with different prices, capabilities, and performance characteristics.

OpenRouter has become popular for access to open-weight models, including systems developed by Chinese AI companies such as DeepSeek and Z.ai. These models have attracted developers partly because they can offer lower costs than proprietary systems from companies such as OpenAI and Anthropic.

Stripe said businesses are increasingly struggling to manage AI costs because models are being released and repriced at a rapid pace. Its interest in OpenRouter is therefore not simply about gaining exposure to AI models, but about controlling the financial infrastructure surrounding their use.

“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 CEO Patrick Collison said.

The logic behind the deal is considered valid. As companies use multiple models for different tasks, selecting the cheapest or most capable model for each request can have a material effect on operating costs. A routing platform can direct a query to different models based on factors such as price, latency, availability, and performance.

That potentially positions OpenRouter as an important layer in the emerging AI software stack. Instead of betting on which individual model will dominate, Stripe is acquiring infrastructure designed to allow businesses to use many models simultaneously.

OpenRouter said the combination would support its goal of creating “a healthy AI ecosystem where many models thrive,” arguing that having multiple competing models reduces the risk that one system becomes the industry default simply because developers are locked into it.

The acquisition also marks a significant expansion of Stripe’s strategy beyond payments. The company has built its valuation primarily around online payment infrastructure, but it has increasingly expanded into adjacent financial and technology services.

Stripe was valued at nearly $160 billion earlier this year. It also strengthened its cryptocurrency business last year through the $1.1 billion acquisition of stablecoin platform Bridge.

OpenRouter gives Stripe a different route into the AI economy. Rather than competing directly with model developers, Stripe would own a platform that helps businesses consume models from competing providers.

The reported valuation also reveals the extraordinary premium investors are placing on AI infrastructure. OpenRouter’s valuation has reportedly jumped from about $1.3 billion to $7.5 billion in less than three months, illustrating how quickly capital is moving toward companies positioned to benefit from the rapid expansion of AI usage.

The deal could ultimately prove more significant than a conventional technology acquisition because it gives Stripe a foothold in the economics of AI inference, the process of running trained models to generate responses for users. Stripe’s bet is that the companies managing the costs of inference, rather than only the companies building the underlying models, could become major beneficiaries of the AI boom.

U.S. government debt has crossed the $40 trillion threshold for the first time, exposing the widening gap between Washington’s ambitions to reduce federal spending and the fiscal realities of an economy in which interest costs, Social Security and healthcare obligations are growing faster than government revenues.

The Treasury Department’s latest daily statement showed total public debt outstanding at $40.047 trillion on Tuesday, comprising $32.266 trillion held by the public and $7.782 trillion in intragovernmental holdings. The milestone came less than five months after the debt crossed $39 trillion, underscoring how rapidly the federal government’s borrowing requirements are expanding.

The increase also provides a stark test of President Donald Trump’s pledge to bring greater discipline to federal spending. One of the most prominent efforts was the defunct Department of Government Efficiency, or DOGE, an initiative led by Elon Musk during the early part of Trump’s second administration and designed to eliminate waste, reduce government payrolls and cut contracts and programmes considered unnecessary.

DOGE initially set extraordinarily ambitious targets. Musk said in early 2025 that he believed the initiative could identify $1 trillion in savings, after initially discussing a $2 trillion reduction in federal spending. By April, he had lowered his expected savings for fiscal 2026 to about $150 billion.

The $40 trillion debt milestone shows how little those efforts have changed the overall trajectory.

More importantly, a recent review by the Government Accountability Office has raised serious questions about the scale of savings claimed by DOGE. The initiative’s so-called “Wall of Receipts” claimed roughly $110 billion in savings, but GAO found that 96% of the reported savings from cancelled grants could not be verified. It also found that more than $27 billion in contracts described as terminated had not actually been cancelled.

This means cutting a federal employee, cancelling a contract or announcing a programme termination does not necessarily translate into an equivalent reduction in federal borrowing. Some savings may occur in future years, some cancellations may be reversed, and some reported reductions may never have represented genuine budget savings in the first place.

The deeper problem is that DOGE was attacking a relatively small part of the federal spending equation.

The U.S. government spends roughly $7 trillion a year, with around 60% going toward mandatory programmes such as Social Security, Medicare, Medicaid and veterans’ benefits. Those programmes are largely driven by statutory eligibility, demographics and healthcare costs. Cutting discretionary agencies and federal payrolls can produce savings, but it cannot by itself resolve a structural deficit of the scale now facing Washington.

That is why the debt has continued rising even as DOGE pursued aggressive reductions.

The fiscal deterioration is becoming increasingly expensive. The federal government is now paying about $1.17 trillion annually to service its debt, according to Treasury data, equivalent to roughly 19% of federal spending in fiscal 2026.

The Congressional Budget Office projects that net interest costs will rise from about $1 trillion in 2026 to $2.1 trillion by 2036, with cumulative interest payments reaching approximately $16.2 trillion over the decade under current-law assumptions.

That creates a particularly dangerous fiscal feedback loop. As the debt stock expands, the Treasury must issue more securities. If interest rates remain elevated, refinancing that debt becomes more expensive. Higher interest costs then enlarge the deficit, requiring still more borrowing.

The bond market is already beginning to price some of this risk.

The yield on 30-year Treasuries recently reached levels not seen since 2007, while a $25 billion 30-year Treasury auction last week cleared at a yield of about 5.22%, the highest borrowing cost at such an auction since 2001.

The significance extends beyond government finance. Treasury yields underpin borrowing costs throughout the U.S. economy. Higher long-term yields can translate into more expensive mortgages, corporate debt and consumer credit, while also reducing the relative attractiveness of riskier assets.

The Treasury responded on Wednesday by announcing that it would at least double the size of its buyback operations for 10- to 30-year Treasuries to $4 billion per operation. The move is intended to improve liquidity and help contain pressure at the long end of the yield curve, although the scale remains small relative to the roughly $30 trillion Treasury market.

The buybacks, however, do not solve the underlying fiscal problem. They can influence market liquidity and the composition of Treasury issuance, but they cannot eliminate the deficit or reduce the government’s long-term spending commitments.

That leaves Washington confronting a much more difficult question: where can sustainable deficit reduction actually come from?

The answer would require decisions involving the largest components of the federal budget, including entitlement programmes and revenues. That means confronting issues that have historically been politically difficult, such as changes to Social Security and Medicare, reductions in other major spending programmes, higher taxes, or some combination of the three.

Trump’s tax and spending policies have added to that challenge. The Congressional Budget Office estimates that the administration’s One Big Beautiful Bill Act will add $4.7 trillion to federal debt.

This creates an obvious tension in the administration’s fiscal strategy. The government pursued spending cuts through DOGE while simultaneously implementing policies that increase the debt trajectory. The arithmetic makes it difficult for reductions in discretionary spending to offset the much larger forces pushing deficits higher.

The history of the past decade illustrates the scale of the problem.

Federal debt stood at about $19.95 trillion when Trump began his first term in January 2017. It has now more than doubled. Trump added about $7.8 trillion during his first presidency, while debt increased by roughly $8.4 trillion during Joe Biden’s presidency. Since Trump returned to office in January 2025, the debt has risen by another $3.8 trillion.

The pandemic accounts for a major portion of the increase, but it is no longer sufficient to explain the trajectory. The emergency spending associated with COVID-19 has ended, yet the federal government continues to run enormous deficits.

That is the central weakness in the argument that waste-cutting alone can restore fiscal balance.

DOGE’s experience illustrates the limits of trying to solve a structural budget problem through administrative efficiency. Eliminating waste is useful and can improve the efficiency of government, but economists say the savings must be measured against a federal budget dominated by entitlement spending, healthcare costs, and interest payments.

Even eliminating every dollar claimed by DOGE would not fundamentally alter the debt trajectory if annual deficits remain measured in trillions of dollars.

The consequences are already spreading into financial markets. Foreign investors, who own nearly one-third of Treasury securities, have reduced their holdings over the past year, meaning more U.S. debt must be absorbed by domestic investors. That can make Treasury markets more sensitive to price and yield movements.

The problem becomes more acute if inflation remains elevated. Higher inflation can keep interest rates higher for longer, increasing the cost of refinancing the government’s debt. Tariffs, geopolitical tensions and higher energy prices could further complicate that environment.

The U.S. still possesses substantial advantages. The dollar remains the world’s dominant reserve currency and Treasury securities remain foundational to the global financial system. Crossing $40 trillion does not mean the United States is suddenly unable to finance itself.

But the margin for fiscal error is narrowing.

The most important lesson from the $40 trillion milestone is therefore not that the United States has reached an arbitrary debt number. It is that years of deficits have reached a point where interest payments themselves are becoming a major driver of future deficits.

DOGE demonstrated that Washington can cut individual programmes, contracts and government jobs. It has not demonstrated that the federal government can reduce its structural deficit.

Economists have warned that until policymakers address the much larger gap between mandatory spending and revenues, the debt will continue to rise regardless of how aggressively government agencies are trimmed.