$60m · Series B · Software · New York City, US
Daytona, a New York City-headquartered provider of execution infrastructure for AI agents, has raised $60m in Series B funding led by FirstMark Growth at a $1bn valuation. The company says the round brings total funding to $91m and includes new strategic investors Qualcomm Ventures, Samsung Next and Okta Ventures, alongside new and returning institutional and individual backers.
Agent workloads turn environments into a usage layer
Daytona gives AI agents isolated computers, browsers and storage in which they can write and run code, use tools and complete workflows. Unlike a conventional cloud instance intended to remain online, the company says an agent environment can start in milliseconds, pause without losing its state, fork into parallel copies and disappear when the task is finished. Its platform spans Linux, Windows, macOS and GPU sandboxes.
That operating pattern creates a different infrastructure bill. One human developer may keep a small number of environments running, while an agent can launch many short jobs and branch them when it tests alternatives. Daytona therefore grows with the number and complexity of agent actions, not only with customer headcount. Fast startup, isolation and state persistence are part of the commercial product because slow or unreliable environments compound across every agent iteration.
The company reports that revenue has grown approximately 25x since January and that average monthly net revenue retention has been about 135%. It also says Writer moved from maintaining more than 20,000 sandboxes a day in-house to provisioning 300,000 Daytona sandboxes in one month. These are company-reported figures, but they illustrate the expansion mechanism: customers may begin with a sandbox requirement and consume much more infrastructure as agents move into production.
Distribution follows production use
Daytona names Meta, Cohere, Writer, Clay, n8n, Postman and Mintlify among customers across model development, training data and agent-enabled products. AfterQuery reportedly replaced an in-house GKE cluster with Daytona in a day. The appeal is not only raw compute; it is avoiding the engineering work required to build isolation, lifecycle management and multi-operating-system support around each agent workload.
The round arrives less than nine months after Daytona's $24m Series A. Existing investors E2VC, Upfront Ventures, FirstMark Capital and Pace Capital returned, while the company is hiring on site in San Francisco across engineering, go-to-market and operations. The next test is whether rapid usage expansion can become durable production demand: as agents receive broader permissions and handle more valuable work, customers will expect the execution layer to remain fast, isolated and dependable across a growing range of environments.
