Every app the agents build is a real web app running on a dedicated cloud server that you configure — the right amount of CPU and memory, an optional GPU, in the region you choose. Pick components, describe the connections, and the agent builds the app around your intent.
Not a shared sandbox. Your app gets its own dedicated machine, which you can resize, move, schedule or switch off.
From a small always-on instance to large multi-core machines for heavy data work.
Attach a GPU for vision, speech and model workloads — object detection, vocal analysis, custom ML.
Run close to your users and your data — Iowa, Sydney, Melbourne or Tokyo.
Start it when you need it, or put it on a schedule. Billed only while it runs.
An interactive, AI-powered workspace — ideal for internal tools and data apps.
A full private or public web app with database and content management — for customer-facing products.