Analysis
Daytona Platforms raised a $60 million Series B at a $1 billion valuation, TechCrunch reported, with FirstMark Growth leading the round and Qualcomm Ventures, Samsung Next and Okta Ventures joining. The New York-based startup says it has now raised $91 million in total.
## What Daytona actually builds Three-year-old Daytona runs a cloud built specifically for long-running AI agents rather than traditional request-response software. The pitch is that agents executing multi-step tasks over minutes or hours need different compute scheduling, state persistence and cost models than the containers most cloud providers sell today. FirstMark also led Daytona's $24 million Series A in February, meaning the firm has now backed the company twice in eight months as its valuation moved to $1 billion.
“FirstMark also led Daytona's $24 million Series A in February, meaning the firm has now backed the company twice in eight months as its valuation moved to $1 billion.”
## The angel list is the real signal What stands out is less the round size than who wrote checks: Stripe's Patrick Collison, Shopify's Tobi Lutke, Perplexity's Aravind Srinivas, Zoom's Eric Yuan, Datadog's Olivier Pomel, Quora's Adam D'Angelo and Mercor's Brendan Foody all participated. That is an unusually deep bench of operator-investors for a $60 million round, and it suggests Daytona's customer base already includes technical buyers who understand agent infrastructure costs firsthand rather than generalist growth investors chasing a hot category.
Daytona's pitch competes directly with the agent-hosting layers AWS, Google Cloud and Microsoft Azure are all racing to build into their core platforms, plus upstarts like Modal and Fly.io that already serve AI workloads. The risk for Daytona is that the hyperscalers have the capital and distribution to commoditize agent-specific infrastructure the moment it proves valuable, the same pattern that squeezed earlier serverless-compute startups once AWS Lambda matured. A $1 billion valuation on $91 million raised also leaves Daytona little room to miss: that is a markup few infrastructure startups sustain if enterprise agent adoption plateaus before the hyperscalers move in.
Still, the speed of FirstMark's follow-on -- a second lead check in under a year -- is the kind of signal institutional investors watch closely heading into 2027, when several of this year's AI-infrastructure bets will need to show real usage rather than just funding announcements.