Daytona, the San Francisco startup building a cloud designed specifically for AI agents, said on Tuesday it has raised a $60 million Series B round at a $1 billion valuation. The round takes the company’s total funding to about $91 million and makes the three-year-old company a unicorn, on the strength of a simple pitch: autonomous agents need computers of their own, and the conventional cloud was not built to give them any.
FirstMark led the round, as it led Daytona’s $24 million Series A before it. Fortune, which first reported the valuation, said roughly $50 million of the Series B came from existing investors.
Daytona’s Series B: Who Backed The Agent Cloud
The investor list reads like a roll call of founder-operators. Individuals joining the round include Stripe’s Patrick Collison, Shopify’s Tobi Lütke, Perplexity’s Aravind Srinivas, Quora’s Adam D’Angelo, Zoom’s Eric Yuan, Datadog’s Olivier Pomel and Mercor’s Brendan Foody. Strategic investors Qualcomm Ventures, Samsung Next and Okta Ventures also took part, alongside Alumni Ventures, Evergreen Fund, Torch Capital, Pace Capital, Upfront Ventures and E2VC.
Daytona was founded in 2023 and is led by chief executive Ivan Burazin. Before the Series B it had raised a pre-seed of about $2 million in 2023 and a $5 million seed in 2024, then the $24 million Series A — a climb from a few million dollars to a $1 billion valuation in roughly three years.
What Daytona’s Agent Cloud Actually Sells
Daytona’s product is an “agent cloud”: secure, isolated environments — sandboxes — in which autonomous agents can run code, use tools and complete work without touching a company’s production systems. The company says the platform combines fast startup, flexible compute, persistent state and granular controls in one infrastructure layer, used for agent training, evaluation and production workloads.
The customer list is the strongest evidence for the pitch. Daytona says it provides infrastructure to Anthropic, OpenAI and NVIDIA, and counts Cursor, Cognition, Meta, Cohere, Mercor, Handshake, Surge AI, SambaNova, Postman, Bolt.new, n8n and Clay among its customers. It also reports growing use for AI training and evaluation, with Snorkel and Trajectory using its infrastructure to cut processing times, and a 25-fold increase in its revenue run rate between January and September.
Burazin framed the safety case in terms any operations manager would recognise: companies are going to manage thousands of digital employees, each one needs a safe place to fail, and Daytona’s job is to make sure one rule-breaking agent does not spill into company-wide chaos.
Where The $60 Million Goes
Daytona said the new funding will expand the agent cloud platform beyond its initial sandbox product, scale its computing infrastructure, and grow its international team, with hiring planned across San Francisco, Croatia and Singapore. The Croatian link is part of the company’s founding story — it was built by a Croatian-founded team before centring on San Francisco.
The round lands in a market where agent infrastructure has become a category of its own. GERoundtable’s survey of the agent platforms to know in 2026 tracks the orchestration layer; Daytona is selling the layer underneath it — the machines the agents run on. The pattern matches other recent agent-economy raises covered here, from Town’s $108 million Series B for team AI assistants to Eka Robotics’ $150 million Series A for physical AI.
Why It Matters
Daytona’s central entity is the agent sandbox, and its attributes tell the wider story: isolation as a safety property, startup speed as a cost property, and a customer roster that already includes the three labs — Anthropic, OpenAI and NVIDIA — whose agents most need somewhere to run. Its neighbours are the agent platforms above it, the hyperscale clouds beside it, and the enterprises that will not let an autonomous agent near production data without a wall around it. A $1 billion valuation for agent infrastructure, three years after founding, is a market signal that the bottleneck in the agent economy is shifting from model quality to the environments agents can be trusted to work in.
Background
Daytona began in 2023 as developer-environment infrastructure — fast, reproducible workspaces for human engineers — before repositioning around AI agents as labs started running autonomous systems at scale. Its Series A in 2025, also led by FirstMark, funded the pivot to agent sandboxes; the Series B announced on October 20, 2026, values that bet at $1 billion.
Source: Fortune via DNYUZ; Tech.eu
0 Comments