Inference-Optimized Compute

Instances optimized for ML inference workloads.

Inference-Optimized Compute

Features

Everything you need to manage your Inference-Optimized Compute infrastructure

Neural processing units

Low latency inference

Model optimization

Batch processing

Auto-scaling

Performance monitoring

Pricing Calculator

Estimate your costs based on your usage

Starting at $0.150 per instance/hour

Cost Calculator

Estimated Cost:$0.15/month