Quick Start

Arkane Cloud offers three ways to work with AI:

• Inference: use hosted models in the playground or through the API.

• Serverless: deploy a model on your chosen GPU configuration with automatic scaling.

• Instances: rent a dedicated CPU or GPU server with full SSH access, and install and run whatever you need.

Create an account to get started.

Try a model

1. Open the Models page and click Try in Playground on a model's card.

2. Type a prompt and send it.

To use a model in your application, follow the inference Getting Started guide.

To deploy a model on your own GPUs that scale automatically, see Deploy your first model.

For full control over your own server, see Set up a CPU or GPU instance.