Deployments

    Deployment Guides

    Deploy from the catalog or configure a model from scratch.

    Deployment paths

    1. Use a published endpoint

      Call a ready production endpoint without deploying it first.

    2. Deploy From Catalog

      Choose a supported model template and deploy it with the provided configuration.

    3. Build From Scratch

      Configure the model, image, runtime parameters, GPU allocation, and environment values yourself.

    InferX Console guide

    The existing step-by-step guide covers login, model creation, vLLM configuration, GPU selection, snapshot creation, testing, and retrieving client values.

    Open the deployment guide →

    Snapshot creation

    After saving a model, InferX schedules snapshot creation. The Console’s Pods view shows progress and the model detail page provides status, logs, and a sample request.