Sandboxes
Start an interactive Polyaxon environment, connect your tools, and automate development and debugging in the same runtime as your workloads.
Start here
What is an AI sandbox? Development environments and execution isolation
Distinguish interactive environments from runtime isolation and define their access.
Continue learning
From notebooks to repeatable ML jobs
Turn interactive experiments into jobs with explicit code, environments, and inputs.
Docker build caching for ML workloads on Kubernetes
Reduce environment rebuilds and startup waits with deliberate build and runtime caches.
Use Notebooks
Use notebooks alongside sandbox, SSH, and terminal access.
How to red team AI agents
Test agent permissions and side effects in a controlled environment.
Compare platforms
Apply the concepts above to a documented platform decision, including where each option fits and when they can coexist.
Polyaxon vs Daytona
Choose between interactive access inside governed workloads and isolated, API-created computers for generated code.
Polyaxon vs E2B
Compare workload-bound interactive environments with managed Firecracker microVM sandboxes.
Polyaxon vs Runloop
Compare a Kubernetes AI platform with isolated Devboxes and agent-specific benchmark workflows.