API Resources

The Datature Vi SDK exposes every platform resource through a single client object. Once you initialize vi.Client, you get access to every resource below, grouped here by the stage of work they belong to. All resources share consistent pagination, error handling, and type hints.

Before You Start

Get started with the Vi SDK →

Data resources

Datasets

List, get, export, download, and delete datasets. The root container for all your training data.

Folders

Organize dataset assets into a browsable virtual tree without moving files.

Assets

Upload, download, list, and delete image and video files within a dataset.

Asset Sources

Inspect where a dataset's assets came from and track ingestion sessions.

Annotation resources

Training resources

Training Projects

Create, list, get, update, and delete training projects. Manage aggregation insights.

Flows

Create, list, get, update, and delete training workflows that define model architecture and hyperparameters.

Runs

Create, list, get, update, kill, and delete training runs.

Models

List, get, and download model checkpoints from completed training runs.

Serving and account resources

Deployments

Manage serving instances, deployment projects, and endpoints, and read live runtime statistics.

Organizations

Get organization info, billing details, user lists, and enabled features.

Plans

Browse the public subscription plan catalog and its feature limits.


For client setup, authentication, pagination, error handling, and other SDK fundamentals, see Getting started with the Vi SDK.


Next steps

Getting Started

Install the SDK, initialize the client, and run your first call.

Vi SDK Inference

Load downloaded models and run predictions on images.

Vi SDK Installation

Install the Vi SDK and configure authentication.


Did this page help you?