--- title: "Daytona Raises $60M Series B at $1B for the Agent Cloud" description: "Daytona raised a $60M Series B led by FirstMark at a $1B valuation. The investment case turns on agent-cloud margins, retention, security and hyperscaler competition." date: 2026-10-21T03:16:00.000Z source: https://venturecapitaltracker.com/daytona-60m-series-b-agent-cloud --- # Daytona Raises $60M Series B at $1B for the Agent Cloud > Daytona raised a $60M Series B led by FirstMark at a $1B valuation. The investment case turns on agent-cloud margins, retention, security and hyperscaler competition. Daytona has raised a **$60 million Series B** led by FirstMark at a reported **$1 billion valuation**, moving the AI-infrastructure company into unicorn territory less than a year after its $24 million Series A. The round brings Daytona’s total disclosed funding to **$91 million**. New strategic investors include Qualcomm Ventures, Samsung NEXT and Okta Ventures. Alumni Ventures, Evergreen Fund and Torch Capital also joined, alongside existing investors FirstMark Capital, Pace Capital, Upfront Ventures and E2VC. The company separately named a group of technology founders and executives as angel investors. This is a confirmed **equity financing**, not debt, a grant or an acquisition. Daytona did not disclose the percentage sold, post-round cash balance, revenue, gross margin or valuation methodology. ## What Daytona is building Daytona describes itself as an “Agent Execution Cloud”: isolated computers that AI agents can create, pause, copy and use to run code, browse, store files and operate software. That positioning matters because the product sits below the visible AI assistant. It is not an inference model or a workflow dashboard. It is the execution layer where an agent performs work. Daytona argues that general-purpose cloud infrastructure was designed for long-lived services, while agents generate large numbers of short, stateful and unpredictable jobs. Its documentation describes Linux containers and virtual machines, Windows environments, macOS access and GPU sandboxes. The company also advertises persistent state, programmatic control through SDKs and APIs, and startup times below 90 milliseconds for its core sandboxes. Those features address a genuine engineering problem. An agent that writes code, tests software or uses a browser needs a controlled environment with permissions, networking and storage. At scale, customers also need isolation, auditability, fast provisioning and a way to destroy environments when work ends. ## The growth claim—and its limits Daytona says revenue increased approximately **25× from January to September 2026**. The company also reports average monthly net revenue retention of about **135%**, meaning its existing customer cohort spent materially more over time. The direction is encouraging, but the disclosure needs context. Daytona did not provide the January revenue base, September run rate, audited figures or gross margin. A 25× increase from a small starting point can still leave a modest business, while infrastructure revenue can carry significant compute, networking and support costs. Monthly net revenue retention is also unusual to compare directly with the annual retention figures commonly reported by public software companies. The metric may reflect rapid workload expansion during an early adoption phase. Investors still need to know how much of that expansion persists, how concentrated usage is among the largest customers and whether pricing covers the cost of increasingly demanding workloads. Company-reported customer examples include Meta, Cohere, Writer, Clay, n8n, Postman and Mintlify. Daytona says Writer moved from an internally operated system handling more than 20,000 daily sandboxes to 300,000 Daytona sandboxes in one month. That is a useful scale reference, but it does not reveal the contract value, margin or long-term retention. ## Why the $1B valuation is difficult to underwrite The $1 billion valuation gives Daytona a strong fundraising signal, but the public information is not sufficient to calculate a revenue multiple. Without annual recurring revenue, gross margin, customer concentration and capital-expenditure data, the valuation is best read as an investor bet on category formation. The thesis is that AI agents will create a new infrastructure layer, and that a specialist provider can capture it before incumbent clouds make agent-native execution a standard feature. The capital requirement could be meaningful. Daytona must support multiple operating systems, storage, networking, security and accelerators while delivering low latency and high availability. GPUs and dedicated macOS hardware add procurement and utilization constraints. If demand is bursty, the company must either carry spare capacity or rely on upstream cloud providers whose costs compress margin. The strongest version of the thesis is not simply that agents need computers. It is that Daytona can pool workloads efficiently, automate the lifecycle better than internal platforms and maintain enough differentiation to earn a durable margin. ## Competitive position Amazon Web Services, Microsoft Azure and Google Cloud already offer virtual machines, containers, serverless compute and managed AI services. Large customers can also build internal sandbox systems. Daytona’s wedge is a lifecycle designed specifically for agents: rapid startup, isolation, persistent state, forking and access to Linux, Windows, macOS and GPUs through a common control plane. If that reduces engineering effort and improves utilization, the product can be valuable even when underlying compute still comes from larger infrastructure providers. The risk is feature compression. Hyperscalers can add faster startup, agent-oriented APIs and security controls. Specialist competitors can target the same sandbox layer. Daytona therefore needs defensibility in reliability, orchestration, workload density, operating-system coverage and integrations—not only in headline provisioning speed. Security is part of the product, not a secondary feature. Agent sandboxes can execute untrusted code, access external services and handle sensitive data. Buyers should examine kernel isolation, secrets handling, network policy, logging, abuse prevention, vulnerability response and evidence from independent audits. ## What the Series B should prove The financing gives Daytona room to scale infrastructure and enterprise support. The most important milestones are measurable: - disclose a revenue range or repeatable growth metric with a clear base; - demonstrate gross margin after compute, storage, networking and support; - show retention across customer cohorts after initial experimentation; - reduce concentration risk among the largest AI companies; - publish reliability, latency and security evidence across operating systems; - prove that usage growth does not require proportionate infrastructure spending. The company has already made a difficult strategic move, shifting from cloud development environments toward agent execution. The $60 million Series B validates investor demand for that thesis. It does not yet prove that the category will support independent infrastructure-scale economics. ## Bottom line Daytona’s **$60 million Series B at a $1 billion valuation** is one of the clearest bets that AI agents need a dedicated execution cloud. The company has credible customer names, rapidly expanding company-reported usage and an investor group that mixes financial and strategic capital. The missing evidence is economic rather than conceptual. Daytona has not disclosed the revenue base behind its 25× growth claim, the margins on multi-platform compute or how concentrated usage is. The next stage is about turning fast sandbox provisioning into a durable, secure and profitable infrastructure business. **By:** [Venture Capital Tracker](https://venturecapitaltracker.com/editorial-policy) **Last updated:** October 21, 2026 **Editorial note:** AI tools assisted with research, structure, or drafting. Venture Capital Tracker retains human editorial responsibility for factual accuracy, relevance, and source quality before publication.