Daytona Platforms announced on October 23 that it raised $60 million in a Series B round at a $1 billion valuation, a bet that the next big cloud will be built for machines, not people. FirstMark Growth led the round, and the Daytona AI agent cloud now has $91 million in total funding behind it.

That’s a unicorn in three years.

Founder Ivan Burazin put the pitch in one sentence. “Agents don’t work like the long-lived services today’s cloud was designed for. They spin up thousands of short, stateful jobs nobody planned in advance, and they need a cloud built for that,” Burazin wrote in a blog post announcing the round.

It’s a simple idea with an expensive implication. An AI agent that writes code, browses the web, or operates business software needs a computer to act on, just as a human worker needs a desk. Give thousands of agents that desk at once, and the old way of renting servers starts to look slow, wasteful, and hard to secure.

Why investors paid $1 billion for the Daytona AI agent cloud

FirstMark isn’t new to this story. The firm also led Daytona’s $24 million Series A in February 2026, and about $50 million of the new round came from existing investors, according to Dealroom. That kind of follow-on money says the early backers liked what they saw in the eight months between rounds.

Then there’s the angel list, which reads like a guest list for the companies actually building agents. Stripe’s Patrick Collison, Shopify’s Tobi Lütke, Perplexity’s Aravind Srinivas, Zoom’s Eric Yuan, Datadog’s Olivier Pomel, Quora’s Adam D’Angelo, and Mercor’s Brendan Foody all joined the round. Qualcomm Ventures, Samsung Next, and Okta Ventures came in as strategic investors.

Founders don’t usually attract that crowd by accident. Daytona says its customers already include Anthropic, OpenAI, Nvidia, Cursor, Cognition, Meta, and Cohere, which use its sandboxes to train and test models. Dealroom also reports that data company Snorkel alone has provisioned more than one million sandboxes on the platform.

Growth numbers from the company point the same way, though they come with the usual caveat: Daytona hasn’t disclosed its actual revenue. It says its revenue run rate grew 25-fold between January and September 2026, averaging 50.9% growth per month, and Pulse2 reports average month-over-month net revenue retention of about 135%. If those figures hold up, customers aren’t just trying the product. They’re spending more on it every month.

What Daytona actually sells

Strip away the funding headlines and Daytona sells sandboxes: isolated virtual machines where an agent can do real work without touching anything sensitive. Each agent gets its own computer, with access to Linux, Windows, macOS, and GPU environments on one platform.

Speed is the selling point. Conventional cloud instances can take 30 seconds to several minutes to spin up, Daytona argues. Its pitch is 60 milliseconds, achieved by keeping pre-configured machine snapshots on physical drives and attaching compute and memory on demand. For a human opening a laptop, that difference barely matters. For an agent that needs to fire off thousands of jobs in a burst, waiting minutes per job kills the whole idea.

Daytona’s environments are also built to pause and resume, fork into parallel copies, and vanish when a task finishes.

Security is the other half of the pitch. An agent running code it wrote itself is a risk if it shares a machine with anything else, and an isolated box gives a bad prompt injection nowhere useful to go.

The cloud giants are moving the same way

Daytona didn’t invent this thesis. On October 22, one day before Daytona’s announcement, Google unveiled Agent Compute Engine at AGNTCon + MCPCon in San Jose. Google’s service gives each agent its own hardware-isolated microVM, holds no credentials inside the sandbox, and pauses idle agents by checkpointing them. Customer preview is still months away, but the service already powers Gemini’s own agents at Google.

AWS moved on the same problem from the open-source side, releasing Infinity, a runtime built around suspending agents while they wait on tools so idle agents cost almost nothing to keep around.

Why this matters

Valuations in AI have a habit of running ahead of revenue, and Daytona’s is no exception: a $1 billion price tag on a company that won’t disclose its revenue is a bet on a trend, not a measurement of a business.

Still, the trend is hard to dismiss: agents are moving from demos into production, and every company deploying them will eventually have to answer where they run.

Daytona just raised $60 million to be that answer. Whether it stays independent, gets bought by a cloud giant, or gets squeezed by one, the question it raised is now the industry’s question: if agents are the new users, who builds their computers?

Frequently asked questions

What is the Daytona AI agent cloud?

It is a cloud platform that gives each AI agent its own isolated computer, with sandboxed Linux, Windows, macOS, and GPU environments, so agents can write code, browse the web, and run software without touching a user’s own machine.

How much did Daytona raise, and at what valuation?

Daytona raised $60 million in a Series B round led by FirstMark Growth at a $1 billion valuation, announced October 23, 2026. The round brings its total funding to $91 million.

Who invested in Daytona’s $1 billion round?

FirstMark Growth led the round, with angel investors including Stripe’s Patrick Collison, Shopify’s Tobi Lütke, Perplexity’s Aravind Srinivas, Zoom’s Eric Yuan, Datadog’s Olivier Pomel, Quora’s Adam D’Angelo, and Mercor’s Brendan Foody, plus Qualcomm Ventures, Samsung Next, and Okta Ventures.

Why do AI agents need their own cloud?

Agents spin up thousands of short, stateful jobs that nobody plans in advance, unlike traditional software that runs around the clock. Daytona argues that work needs computers that start in milliseconds, hold state while paused, and disappear when the job is done.

Sources: TechCrunch, Dealroom, Pulse2, Fortune (via Tech Startups).