SDK

Lightning AI provides a powerful set of tools across the SDK, CLI, and API to help you programmatically manage resources like Studios, Jobs, and Deployments. Whether you're automating workflows or integrating Lightning AI into your existing stack, this section covers everything you need to interact with the platform.

Studios

Studios are interactive, persistent environments for developing and debugging machine learning workflows. Use the SDK, CLI, and API to create, manage, and access Studios directly from your code, terminal, or external systems.

Key features:

• Launch and terminate Studios via CLI or Python
• Programmatically control Studio resources
• Integrate Studios with data workflows and training pipelines

Jobs

Jobs allow you to run code in the cloud at scale, from training models to data processing tasks. With the SDK, CLI, and API, you can create, monitor, and manage jobs seamlessly across environments.

Key features:

• Submit and configure cloud jobs
• Track job progress, logs, and status
• Automate job execution in pipelines

Deployments

Deployments make it easy to turn models or apps into scalable, production-ready services. The SDK, CLI, and API let you deploy, update, and control your endpoints for real-time inference or app hosting.

Key features:

• Programmatically launch and manage deployments
• Push updates to models or app code
• Monitor deployed services and scale resources

Sandboxes

Sandboxes are isolated, on-demand Linux environments you can create and control programmatically. Each sandbox gets its own filesystem, network stack, and dedicated compute — ideal for running untrusted code, agent workloads, or ephemeral build and test environments.

Key features:

• Spin up and tear down ephemeral environments via Python or TypeScript
• Run commands, manage files, and open interactive PTY sessions
• Capture and restore filesystem snapshots