Amazon Bedrock is a fully managed service providing access to high-performing foundation models via a single API, offering security, privacy, and responsible AI capabilities.
Pricing Structure
Charges apply for both model inference and customization.
Inference Pricing Plans
- On-Demand and Batch: A pay-as-you-go model with no time-based commitments.
- Provisioned Throughput: Allows provisioning of sufficient throughput to meet application performance requirements in exchange for a time-based commitment.
On-Demand Mode
- Charges are based on actual usage.
- No time-based commitments.
- Supports cross-region inference for some models without additional charges.
- Pricing is determined by the source region where the request is made.
Provisioned Throughput Mode
- Requires purchasing model units for a specific base or custom model.
- Best suited for large, consistent inference workloads requiring guaranteed throughput.
Please refer to the website Amazon Bedrock to read more about this.