Daytona, a provider of agent-specific cloud infrastructure, has announced a US$60 million Series B round at a US$1 billion valuation. The funding was led by FirstMark Growth and lifts the company's total raised capital to US$91 million. The announcement marks a shift from early contrarian positioning to broad consensus around the need for dedicated execution environments for AI agents.
Building the Agent Execution Cloud
Daytona describes the concept as an Agent Execution Cloud, focused on the execution layer that agents require rather than inference or observability. The company built its platform around computers, browsers, storage, and related tools that let agents perform real work. Its core innovation makes a computer for every agent practical through instant startup, persistent state, and native support for Git, code execution, and safe operations.
A Different Cloud Model
Traditional cloud instances take minutes to boot and run until someone shuts them down. Agent computers must start in milliseconds, pause with intact state, fork into parallel copies, and disappear when a job ends. Daytona offers Linux, Windows, macOS, and GPU sandboxes on one platform, allowing agents to test native Mac apps, drive Windows desktops, optimize CUDA kernels, or work in 3D CAD software.
Investor Confidence and Strategic Backing
The Series B round includes individual investors such as Patrick Collison of Stripe, Tobi Lütke of Shopify, Aravind Srinivas of Perplexity, Eric Yuan of Zoom, Olivier Pomel of Datadog, Adam D'Angelo of Quora, and Brendan Foody of Mercor. Qualcomm Ventures, Samsung Next, and Okta Ventures joined as strategic investors, with additional participation from new and existing backers. Existing investors E2VC, Upfront Ventures, FirstMark Capital, and Pace Capital also returned for this round.
Rapid Commercial Traction
Daytona reported that revenue has grown approximately 25 times since January, with average month-over-month net revenue retention of approximately 135%. Customers from six months ago are, in aggregate, spending nearly three times as much today, according to the company. This retention pattern indicates that existing customers are expanding their use of Daytona's infrastructure as agent workloads scale.
Customers Across the Agent Stack
The company's customer base spans model builders, data and environment teams, and product teams putting agents into everyday applications. Named customers include Meta, Cohere, Handshake, Snorkel, Surge AI, Turing, AfterQuery, Trajectory, Writer, Clay, n8n, Postman, Mintlify, and bolt.new. Writer moved from an in-house sandbox platform to Daytona after surpassing 20,000 sandboxes a day, then provisioned 300,000 sandboxes in a single month.
Platform Momentum and Future Plans
Daytona believes the next infrastructure shift will move away from human-centric cloud primitives toward agent-native ones. The company says agents do not work like long-lived services and need a cloud built for thousands of short, stateful jobs that were never planned in advance. Daytona is hiring across engineering, go-to-market, and operations at its San Francisco headquarters as it continues building the infrastructure agents will run on.
The funding round positions Daytona as a key infrastructure provider for the growing agentic economy. With a US$1 billion valuation, strategic backers, and rapid commercial traction, the company is moving from early experimentation to production-scale deployment. Its focus on agent-specific execution environments suggests that the market is now treating agent infrastructure as a distinct and essential cloud category.