Amazon Bedrock model pricing
Estimate the token cost of any Amazon Nova model on Bedrock, before you build.
Start here when you just need to know what raw model inference will cost. Every other architecture on this page adds services on top of this number.
Estimate your workload
Model choice is usually the biggest cost lever. Nova Micro/Lite are ~10–70× cheaper than Premier.
Results update automatically as you type.
- Per request
- $0.00960
- Per day
- $9.60
- Per month
- $288.00
- Per year
- $3,456.00
- Model
- Amazon Nova Pro
Frequently asked questions
Which Nova model is cheapest?+
Nova Micro, then Lite, then Pro, then Premier. Micro and Lite handle most extraction and summarisation tasks; reserve Pro/Premier for hard reasoning.
When does batch pricing apply?+
Batch (asynchronous) inference is ~50% cheaper but not real-time. Use it for bulk document processing where a few minutes of latency is fine.
Other AWS architectures
AgentCore and managed-service rates last verified 14 July 2026 (us-east-1). Bedrock model prices last reviewed 28 July 2026. Estimates for planning, not a quote.