NEW YORK — Ramp announced on October 15, 2026, that it had raised $1.85 billion in primary financing at a $60 billion valuation. The company's announcement accompanies a push into tools that help businesses use AI, measure its cost and share the work it produces.
The New York connection goes beyond the announcement's dateline. Ramp identifies the city as its headquarters in its recruiting materials, with roles spanning applied science and software engineering. Its expansion offers a local example of a financial technology business moving into the infrastructure around enterprise AI.
A larger round for a broader business
Reuters' October 15 report confirms Dragoneer Investment Group led the financing, with Thrive Capital, Founders Fund, Sands Capital and Avenir Growth as co-leads. The transaction is described as primary financing, meaning capital raised for the business, rather than simply a sale of existing shareholders' holdings.
The Reuters coverage puts the new valuation alongside June's $44 billion valuation and $750 million round. That makes the latest valuation roughly 36% higher, a comparison calculated from the two disclosed figures.
Reuters also reported that Ramp, founded in 2019, said it now serves more than 85,000 businesses and has raised approximately $5 billion in total equity financing. The progression gives the company more capital as it reaches beyond its corporate card and expense management roots.
What the revenue figure means
In its October 1 snapshot, Ramp reports more than $2 billion in annualized revenue and positive cash flow. Its funding release defines revenue as including interchange, subscriptions, interest and other revenue, excluding customer incentives. An annualized measure expresses a revenue pace over a year; it does not establish that an entire year at that level has been completed, and cash flow is a different measure from net profit.
Those distinctions matter when assessing the scale of a business that combines financial services with software. TLVNY's reading is that the product expansion deserves attention alongside the financing: Ramp is trying to connect the purchase of AI services, the work people do with them and the finance team's view of the bill.
Making AI spending visible
Ramp's AI Spend Management page describes a dashboard that combines spending across providers and breaks it down by model, user, project and API key. The company says teams can assign usage to an owner or department, investigate spikes and receive weekly explanations of changes. It also offers standalone access without a Ramp corporate card.
The distinction between visibility and enforcement is useful here. The page describes thresholds that notify people when spending reaches a limit. It separately marks custom approval workflows and provider-invoice reconciliation as forthcoming features. A budget alert should therefore not be understood as a guarantee that every AI request will automatically stop at a spending ceiling.
For finance and engineering teams, the practical question is whether a bill can be traced to the person, project or workload responsible for it. That is a more specific objective than simply knowing the organization's total AI expenditure.
Choosing which model handles the work
Ramp's AI expansion also reaches into how software calls models. Its August 19 Router launch post describes a single interface for accessing models, evaluating alternatives and directing requests between them. This is earlier product context for the October financing, rather than a launch that happened this week.
According to that post, teams can compare candidate models using real workloads while their current model continues serving users. The system exposes information such as tokens, latency, cost and fallback attempts for individual requests. Ramp presents the trade-off as choosing an appropriate balance of quality, reliability, speed and expense.
The commercial idea is straightforward: controlling an AI budget can involve changing how work is processed, as well as counting what has already been spent. Product descriptions establish Ramp's intended approach; the result for a particular business depends on its workloads and evaluation criteria.
Glass adds a shared workspace
On October 13, co-founder and co-CEO Eric Glyman introduced Glass, an AI workspace designed to retain and reuse company knowledge. His launch post describes shared sessions, reusable skills and connections to workplace information from services including Linear, Notion, Google, Slack and Gong.
Glyman says the workspace lets colleagues join a session and provides project-level cost meters and limits. He also says requests run through Ramp Router. The page invites businesses to request early access and says the first organizations are being onboarded personally, so this remains a product rollout rather than evidence of universal availability.
Glass extends the story from financial controls into employees' daily work. Its premise is that a useful workflow or instruction developed by one person can become something colleagues build on. Whether that produces durable improvements will depend on how well the shared knowledge is maintained and used.
What to watch from New York
For New York's technology community, Ramp's expansion illustrates a route into enterprise AI through existing business operations. Its tools address several connected questions: what a company pays for, which model performs a task and how useful work can be reused.
The next evidence to watch is Glass adoption beyond early access, the delivery of features still on the roadmap and results that customers can substantiate. The October 15 financing establishes the scale of investors' commitment. The product rollout will show how much of that ambition becomes part of the everyday work of finance and engineering teams.
