Glossary

Less than 1 minute

Key terms and definitions used throughout the Nosana documentation.

A

Using a trained AI model to generate outputs or predictions from new input data.

Authentication token used to access the Nosana API via the SDK or direct HTTP requests.

GPU market segment for new or unvalidated hosts to start, test setups, and prove reliability.

A lightweight, isolated environment that packages your app code and dependencies (e.g. a Docker image).

Usage-based currency consumed when running deployments and jobs on Nosana, claimable from the Dashboard.

The Nosana web UI where you manage API keys, credits, deployments, and monitor workloads.

A long-lived object that describes what to run and how to run it, creating jobs over time.

A pool of GPU resources with shared characteristics where deployments are scheduled.

An individual GPU machine in a GPU market that actually runs your jobs.

A concrete execution of your workload created from a deployment’s job definition.

The JSON specification describing image, commands, and resources for each job.

GPU market segment for validated, high-performing GPU providers suitable for critical workloads.

The TypeScript Software Development Kit for interacting with Nosana.

This glossary is continuously updated. If you find a term that needs clarification, please let us know.