Daytona, which sells isolated computing environments for AI agents to run code, browse the web and operate software, announced on October 20 that it has raised a $60 million Series B at a $1 billion valuation. Fortune, which reported the round first, and other coverage say FirstMark led it.
The round
According to the reports, roughly $50 million of the round came from existing investors, and Daytona’s total funding now stands at about $91 million. Angel investors named in coverage include Stripe’s Patrick Collison, Shopify’s Tobi Lütke, Datadog’s Olivier Pomel, Perplexity’s Aravind Srinivas and Zoom’s Eric Yuan, alongside strategic investors such as Qualcomm Ventures, Samsung Next and Okta Ventures. We did not find a stated plan for how the money will be spent.
What Daytona sells
Daytona provides sandboxes, which it describes as programmatic, composable computers for agents. Its agent execution cloud offers Linux, Windows, macOS and GPU environments. CEO Ivan Burazin argues to Fortune that conventional cloud platforms were designed around human developers and long-running applications, whereas agents need many short-lived machines created on demand. One outlet reports the company can start a sandbox in about 60 milliseconds by pre-staging snapshots; that figure comes from the company and we have not seen independent benchmarks.
Secondary coverage also reports that Daytona’s revenue has grown about 25 times since January, that net revenue retention averages around 135% a month, and that customers include Anthropic, OpenAI, Nvidia and Meta. These figures are company-supplied and appeared in summaries rather than in the Fortune excerpts we could review, so treat them as unconfirmed.
Why it matters
As coding and browsing agents move from demos to production, the layer that actually executes their work, including untrusted code and long browser sessions, has become its own market. Daytona’s valuation suggests investors see that sandbox layer as durable infrastructure rather than a feature of individual agent products. Buyers evaluating agents should note that the underlying execution environment affects speed, isolation and cost, even when vendors do not name it.