Daytona spins up a single sandbox in 60 milliseconds including network latency, and can launch 50,000 concurrent sandboxes in 75 seconds.

“Time to spin up one is 60 milliseconds with network latency.”

Daytona: an agent sandbox infrastructure provider reporting 74% month-over-month growth, 850,000 daily runs, and sub-100ms spin-up times, positioning itself at the intersection of agent compute and cloud infrastructure.

“Time to spin up one is 60 milliseconds with network latency.”

Utilization on Daytona infrastructure for spiky eval and RL workloads is only 15%, illustrating the mismatch between peak provisioning and actual consumption.

“Utilization is 15% 15.”

Agent infrastructure requires fundamentally different design from developer infrastructure, contrary to widespread early assumptions that the two are interchangeable.

“Most people thought it was the same infrastructure for humans and agents. We understood a quarter ago it's not.”

Ivan Burazin on how quickly the industry recognized that agent infrastructure is not the same as developer infrastructure.

“Most people thought it was the same infrastructure for humans and agents. We understood a quarter ago it's not.”

Agent compute workloads are highly spiky with very low average utilization, a usage pattern not seen with human developer workloads, which forces infrastructure providers to overprovision dramatically.

“Usage is super spiky and this is something that has not happened before that you have these types of like it was always so it's quite interesting use case and problem solve.”

Open sourcing Daytona's code accelerates cloud product adoption because AI agents consume the repository as context when integrating the tool, making the codebase itself a distribution mechanism.

“It's actually more helped in the consumption of our cloud product than actually transferring people over. The reason is you can actually you send the repository to your agent when you're integrating Daytona and it just has more context.”

The AI infrastructure market as a whole is growing at roughly 40% month-over-month, with broad participation across providers.

“The entire infrastructure market is growing 40% plus or minus month over month everyone is growing 40% month.”

CPUs, not GPUs, will become the primary bottleneck for agent compute as agent workloads scale.

“CPUs will now be a bottleneck because it will be the constraint. You won't be able to grow or we won't be able to have enough of these because there won't be enough CPUs to basically do that.”