Architecture
Runta is an execution layer for AI agents. It provides scalable, governed runtimes with strong control over state, access, credentials, and execution. Try it now.
Features
Section titled “Features” Runtime Basics Create, execute commands in, resize, pause, resume, stop, and delete runtimes.
Memory Auto Scaling Start with a smaller memory request and grow automatically up to a configured limit.
Files Copy, upload, download, read, and write files in a runtime.
Egress Restrict outbound access with allowed or denied host policies.
Checkpoints Capture runtime state and restore it into new runtimes.
Secret Stubs Store tenant secrets and inject them into matching outbound requests.
Publish a Service Expose a runtime service over HTTPS through Runta ingress.
Auto Suspend and Wake-Up Suspend idle runtimes and wake published runtimes when traffic returns.
Reference
Section titled “Reference” Harbor Use Harbor workflows with Runta runtime isolation.
OpenAI Agents SDK Run OpenAI Agents SDK sandboxes on Runta runtimes.
CLI Reference Commands, flags, configuration, lifecycle, secrets, egress, checkpoints, and TLS.
SDK Reference Python and TypeScript interfaces for runtimes, files, egress, secrets, and checkpoints.
REST API Reference Generated API docs for runtimes, exec, checkpoints, files, egress, and secrets.
Pricing Usage-based pricing for compute, memory, and storage.