Daytona raises $60M Series B at $1B valuation for its Agent Cloud

Funding and investors

According to Daytona, the round is led by FirstMark Growth and values the company at $1B. Total funding now stands at $91M. Several well-known founders took part as angels: Patrick Collison (Stripe), Tobi Lütke (Shopify), Aravind Srinivas (Perplexity), Eric Yuan (Zoom), Olivier Pomel (Datadog), Adam D'Angelo (Quora) and Brendan Foody (Mercor).

Qualcomm Ventures, Samsung Next and Okta Ventures joined as strategic investors. Other new investors include Evergreen Fund, Alumni Ventures and Torch Capital. Existing backers returned as well: E2VC in its fourth round with the company, Upfront Ventures in its third, and FirstMark Capital and Pace Capital in their second.

What the Agent Cloud is

CEO and co-founder Ivan Burazin writes that Daytona has long held that agents are "digital humans" that need a computer in every configuration a human would. He says this was a contrarian view at first, but that long-running agents went mainstream this year and agent-specific infrastructure became consensus.

Daytona describes the Agent Cloud as an execution layer: computers, browsers, storage and anything else an agent needs to do its job. It is not inference, observability or management dashboards. The company argues that agents spin up thousands of short, stateful jobs nobody planned in advance, unlike the long-lived services today's cloud was designed for.

In its view, an agent's computer differs from a traditional cloud instance. An instance takes minutes to boot and runs until shut down. An agent's computer, Daytona says, has to start in milliseconds, pause with its state intact, fork into parallel copies and disappear when the job is done.

Daytona also says it is the only provider offering Linux, Windows, macOS and GPU sandboxes on one platform. Its examples include optimizing a CUDA kernel, testing a native Mac app, driving a Windows desktop or working in 3D CAD software. The company adds that it is already building further tools agents will need.

Growth and customers

Daytona reports that its revenue has grown approximately 25x since January, with average month-over-month net revenue retention of approximately 135%. The company says customers it had six months ago are, in aggregate, spending nearly three times as much today. It compares this to Snowflake's net revenue retention at IPO of around 160% per year, noting that Daytona averages close to that figure every month.

Named customers include model builders Meta and Cohere. Data and training-environment teams such as Handshake, Snorkel, Surge AI, Turing, AfterQuery and Trajectory also use the platform. Product companies include Writer, Clay, n8n, Postman, Mintlify and bolt.new.

Two examples stand out. Writer ran its own sandbox platform until it passed 20,000 sandboxes a day, when keeping it running became a full-time job. On Daytona, it provisioned 300,000 sandboxes in a single month. AfterQuery replaced its in-house GKE cluster with Daytona in a day. Sam Jung of AfterQuery says: "As a frontier applied AI research lab, the infrastructure we need often does not exist. Daytona broke that trend."

What it means

The announcement points to growing demand for infrastructure built specifically for agents rather than adapted from general-purpose cloud services. Daytona's own account is that teams often start by building sandbox systems themselves and then hit operational limits as usage grows. Its pitch is that they can hand that work to a dedicated platform.

Daytona says it is hiring on site in San Francisco across engineering, go-to-market and operations. The figures on growth and retention are the company's own claims and were not independently verified in the announcement.