Everything you need to deploy AI workloads on Apple Silicon
Get up and running in under 5 minutes. Install the SDK, authenticate, and run your first job.
Get started →Complete reference for the MetalCloud Python SDK. Job submission, streaming, and more.
View reference →Direct API access for any language. OpenAPI spec, authentication, and endpoints.
Explore API →Pre-configured models ready to deploy. Llama, Mistral, DeepSeek, and more.
Browse models →Using Apple's MLX framework on MetalCloud. Optimizations, examples, and best practices.
Learn MLX →Understanding your bill, usage tracking, and cost optimization strategies.
View billing →Frequently accessed documentation
Run inference in just a few lines of code
import metalcloud # Initialize client client = metalcloud.Client() # Run inference on Llama 70B response = client.inference( model="meta-llama/Llama-3.3-70B", prompt="Explain unified memory in Apple Silicon", max_tokens=500 ) print(response.text)