Daytona Hits $1B Unicorn Valuation with $60M Series B to Build the Agent Cloud
Daytona Platforms, a New York-based startup, announced this week that it had raised a $60 million Series B round at a $1 billion valuation, led by FirstMark Growth. The round brings the company’s total funding to approximately $91 million. The investor roster reads like a roll call of Silicon Valley’s founding generation: Patrick Collison (Stripe), Tobi Lütke (Shopify), Aravind Srinivas (Perplexity), Eric Yuan (Zoom), Olivier Pomel (Datadog), Adam D’Angelo (Quora and OpenAI board member), and Brendan Foody (Mercor). Strategic investors included Qualcomm Ventures, Samsung Next, and Okta Ventures.
Founded in 2023, the San Francisco-headquartered company is building cloud infrastructure designed specifically for AI agents — software entities that execute tasks autonomously, often requiring full computing environments to write code, install software, browse the web, and interact with APIs.
Why the existing cloud doesn’t work for agents
Daytona’s core argument is that today’s cloud infrastructure was designed for a fundamentally different workload. Traditional cloud services — web servers, REST APIs, microservices — are long-lived, predictable, and relatively stable in their behavior. An AI agent, by contrast, operates in bursts: it might spin up an entire development environment from scratch, install dependencies, compile code, run tests, open a browser window, navigate a web interface, and then shut down the moment its task is complete.
"Agents don’t work like the long-lived services today’s cloud was designed for," wrote CEO and co-founder Ivan Burazin in a blog post announcing the round. "They spin up thousands of short, stateful jobs nobody planned in advance, and they need a cloud built for that."
Daytona’s platform provides what it calls "composable computers" — isolated runtime environments with dedicated kernels, filesystems, network stacks, and allocated vCPUs, RAM, and disk. Each sandbox boots in under 90 milliseconds and can run on Linux, Windows, macOS, or with GPU acceleration (NVIDIA or AMD). State persists across sessions: an agent can pause a task, resume it later, fork it into parallel copies, and delete it when finished.
The agent cloud ecosystem
Daytona’s thesis fits into a broader shift in the AI industry. As autonomous agents become more capable — moving beyond simple chat interfaces to execute complex, multi-step workflows — they require infrastructure that can support that behavior at scale. Anthropic, OpenAI, Google, and others are all investing heavily in agents that don’t just answer prompts but perform real work on behalf of users.
Daytona started as a tool for developers who wanted to run AI agents in isolated environments. The product has since evolved into a full execution platform serving everything from startups to large engineering teams. Customers include Handshake, Snorkel AI, Surge AI, Turing, and AfterQuery, which use the platform to train and evaluate models and deploy agents in production.
The participation of strategic investors like Qualcomm Ventures and Samsung Next is noteworthy: it signals that the agent infrastructure play is being taken seriously not just by software investors but by hardware companies looking to build dedicated silicon or accelerators for agent workloads.
What’s next
With $60 million in the bank and a unicorn valuation, Daytona plans to expand its platform beyond its initial sandbox product, scale its computing infrastructure, and grow its international presence. The company has offices in San Francisco, Croatia, and Singapore and intends to recruit talent across all three.
The real test, however, will come in the months ahead. Can Daytona convert the excitement around agentic AI into durable, production-grade demand? As agents receive broader permissions and handle increasingly valuable work, enterprises will need the execution layer to remain fast, isolated, and reliable across a growing range of environments. If Daytona can demonstrate that its platform is not just a convenience but a necessity — the equivalent of giving every AI agent its own dedicated computer — the agent cloud could become one of the defining infrastructure categories of the next decade.
Sources: TechCrunch, Fortune, EU.VC
✓ Independent sources cross-checked and verified before publishing
