Daytona announced a $60 million Series B at a $1 billion valuation on Oct. 20, 2026, adding capital for infrastructure that lets AI agents execute tasks in computing environments. The financing brings its total funding to $91 million, according to the company.
This Rapid News Brief separates the financing and announced platform capabilities from broader claims about demand and security.
Infrastructure beneath the agent
CEO and co-founder Ivan Burazin’s announcement identifies FirstMark Growth as lead investor. Qualcomm Ventures, Samsung Next and Okta Ventures joined as strategic investors, alongside other new and existing backers.
Daytona describes its product as an execution layer providing computers, browsers and storage, rather than an inference service or a dashboard for supervising agents. The company says its environments can preserve state, pause and resume work, and be copied for parallel tasks.
The operational proposition is to supply the place where an agent works, not simply the model that decides what to do. A buyer evaluating that proposition would need to consider the execution environment separately from its choice of model and its application’s approval rules.
A rebuilt business, with expansion ahead
Tech Funding News reported that Daytona was founded in 2023 by Burazin, chief technology officer Vedran Jukić and chief architect Goran Draganić. Its initial focus was automating development environments for enterprise software teams.
The publication places the new round after a $24 million Series A in February, a $5 million seed round in 2024 and a $2 million pre-seed in 2023. Its coverage describes the company’s shift toward infrastructure for agents rather than its original human-developer product.
Daytona’s announcement says it is hiring across engineering, go-to-market and operations in San Francisco. Those expansion plans do not establish a completed increase in capacity or staffing.
For enterprise teams, the funding is a supplier-development signal. It is not independent proof of faster task completion or stronger isolation in a specific deployment. Those questions still require testing against the application’s actual tools, credentials and workload.