VAST Data has closed a $1 billion Series F financing round at a $30 billion valuation, more than tripling its $9.1 billion valuation from 2023. The round was led by Drive Capital and co-led by Access Industries, with participation from existing investors including Nvidia, Fidelity Management and Research, and NEA.
The New York-based company builds what it calls an AI Operating System, a software infrastructure layer that sits between hardware and AI models and manages how data is stored, moved, and retrieved at scale. As a result, its platform currently supports AI environments spanning millions of GPUs in production.
Why the VAST Data Series F Signals a Shift in AI Infrastructure Investment
The raise reflects a shift in where capital is moving. For two years, funding concentrated at the model layer. Now investors are backing the infrastructure beneath it — specifically the companies that determine how fast data can reach compute at scale.
Training and running AI models at the scale that hyperscalers and neoclouds now operate requires managing multi-exabyte datasets with speed and reliability that older distributed storage architectures cannot provide. VAST’s disaggregated shared-everything architecture addresses exactly this problem. As Nvidia CEO Jensen Huang put it in a recent video accompanying the announcement, without software like VAST’s, even the fastest AI processors face severe data bottlenecks.
The company reported more than $4 billion in cumulative bookings and over $500 million in committed annual recurring revenue at the time of closing. VAST has been cash-flow positive for several years and did not raise the round out of operational necessity. CEO Renen Hallak said inbound investor demand entirely drove the size of the round.
Who Is Already Using It and What Comes Next
Customers include CoreWeave, xAI, Mistral AI, the US Air Force, JPMorganChase, and Cursor. Contracts typically run five to seven years. Once organisations integrate VAST’s platform into their AI workflows, switching costs are high and churn is low.
Mistral AI co-founder and CTO Timothée Lacroix described the platform as essential for managing and scaling the massive datasets required to train frontier models at speed. That kind of endorsement from a frontier lab signals that VAST has moved from a storage vendor to a core piece of the AI stack.
Hallak said the company is targeting IPO-readiness by the end of 2026, though no formal timeline has been set. Primary proceeds will fund geographic expansion into Latin America, India, and China, as well as strategic acquisitions that extend VAST’s technology footprint. As we have covered previously, the unit economics of neocloud infrastructure increasingly depend on how efficiently data moves between storage and compute. VAST’s $30 billion valuation is the market’s answer to how much that problem is worth solving.
