Daytona: US$60M at a US$1B Valuation to Give Every AI Agent Its Own Computer
Daytona, a San Francisco-based startup founded by three Croatian engineers, raised US$60 million in a Series B round at a US$1 billion valuation, led by FirstMark Growth. The deal was first reported by Fortune on October 20, 2026. About US$50 million came from existing investors, according to the company. Qualcomm Ventures, Samsung Next and Okta Ventures joined, along 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. Daytona has now raised about US$91 million, including a US$24 million Series A led by FirstMark in February. The company was founded in 2023 by CEO Ivan Burazin, CTO Vedran Jukić and chief architect Goran Draganić. Burazin previously co-founded Codeanywhere and was chief developer experience officer at Infobip.
Daytona gives each AI agent a "sandbox," an isolated virtual computer where it can run code, browse the web or operate software without affecting other systems. The company says a sandbox starts in 60 milliseconds, compared with 30 seconds to several minutes on traditional clouds, and can be paused, resumed or copied so an agent can test several approaches at once. It supports Linux, Windows, macOS and GPU environments. Daytona says it supplies infrastructure to Anthropic, OpenAI, Nvidia, Cursor and Cognition, and counts Meta, Cohere and Mercor among its customers. Its revenue run rate grew about 25 times between January and September, averaging 50.9% growth a month, though the company has not disclosed the base. Most of the new money will go to sales hires, plus some engineers, with hiring planned in San Francisco, Croatia and Singapore.
Market Context
Agent infrastructure has become one of the most heavily funded corners of AI. On October 14, Modal, which also sells sandboxes, raised US$750 million at a US$15.75 billion valuation, more than triple its valuation in May. Temporal, which keeps long-running software processes going when parts fail, raised US$550 million at US$12.55 billion in September. Open-source sandbox provider E2B raised a US$21 million Series A in 2025. Gartner forecasts enterprise spending on AI agents and assistants will top US$23 billion in 2030.
The bet also touches chips. Unlike model training, which runs on GPUs, agent work leans on standard server processors (CPUs). AMD raised its forecast for server CPU market growth to more than 35% a year, reaching over US$120 billion by 2030, citing agentic AI. Arm estimates agent-focused data centers could need four times as many CPU cores per gigawatt as traditional ones. Burazin told Fortune he expects "a CPU crunch."
Key Signal
"The inference clouds is where the agent thinks, and our cloud is where the agent works." Ivan Burazin, co-founder and CEO, Daytona (via Fortune)
Regional Relevance
For the United States, Daytona is part of a new layer of cloud computing being built specifically for AI agents rather than for human-run applications. If agents become routine "digital employees," as Burazin argues, the companies that host them could capture a share of cloud spending that today goes to Amazon, Microsoft and Google. The investor list reflects that: Qualcomm is exposed to the CPU demand agents create, and Okta, an identity and security company, is exposed to the question of how to keep agents from causing damage when they misbehave.
For Croatia and Central and Eastern Europe, Daytona adds to a growing line of globally scaled companies with Croatian founders, after messaging platform Infobip, where Burazin worked, and electric hypercar maker Rimac. Although headquartered in San Francisco, Daytona plans to keep hiring in Croatia, which helps retain senior engineering talent in a region that often loses it to Western Europe and the U.S.
The Other Side
How big is the business behind the US$1 billion price tag? A 25-fold jump sounds dramatic, but Daytona has not disclosed revenue, so the starting point could be small. Tech Funding News noted that rivals report run rates large enough to calculate valuation multiples (Modal above US$1 billion, Temporal above US$250 million), while Daytona's cannot be assessed.
What happens if the biggest customers build their own? Daytona's growth leans on a handful of AI labs and coding companies. Firms like OpenAI and Anthropic have the resources to bring sandbox infrastructure in-house, and the large cloud providers already rent virtual machines. Burazin argues rebuilding for agents is hard for them, but that remains to be proven.
Can it secure enough hardware to keep growing? Burazin himself warns of a CPU shortage, and Intel reportedly told some customers that lead times for certain server chips had stretched to six months. A company whose product is fast-starting machines depends on getting enough of them at a reasonable cost.
Sources & Transparency
- Agent cloud startup Daytona hits $1B valuation, backed by FirstMark and a who's who of angels
- Exclusive: Daytona, a startup building the 'agentic cloud,' valued at $1 billion in latest funding round
- Daytona hits $1B valuation with $60M Series B for AI agent cloud
- Daytona raises $60M at $1B valuation to build a cloud for AI agents
- Daytona Raises US$60 Million Series B at US$1 Billion Valuation