Detailed Pricing for Specific Models on Amazon Bedrock

Amazon Bedrock Pricing Structure

Amazon Bedrock primarily charges for running inferences on third-party foundation models. The costs are determined by the volume of input and output tokens processed by these models. Additionally, for businesses requiring dedicated, faster processing capabilities, Amazon offers 'provisioned throughput', which is a reserved capacity ensuring consistent processing speed at an additional charge. The pricing for each model varies depending on the chosen model from various AI companies​​.

1. AI21 Labs Models:
2. Jurassic-2 Mid: Priced at $0.0125 per 1000 input and output tokens.
3. Jurassic-2 Ultra: Costs $0.0188 per 1000 input and output tokens.
4. Note: Model customization and Provisioned Throughput are not supported for AI21 Lab models on Amazon Bedrock​​.
5. Amazon Titan Models:
6. Titan Text – Lite: Charges $0.0003 per 1000 input tokens and $0.0004 per 1000 output tokens.
7. Titan Text – Express: Priced at $0.0013 per 1000 input tokens and $0.0017 per 1000 output tokens.
8. Titan Embeddings: Costs $0.0001 per 1000 input tokens.
9. Provisioned Throughput pricing ranges from $5.10 to $20.50 per hour, depending on the model and commitment period​​.
10. Anthropic AI Models:
11. Claude Instant: On-demand pricing is $0.00163 per 1000 input tokens and $0.00551 per 1000 output tokens.
12. Claude: Priced at $0.01102 per 1000 input tokens and $0.03268 per 1000 output tokens.
13. Provisioned Throughput for Claude models ranges from $22.00 to $63.00 per hour, depending on the model and commitment period​​.
14. Cohere Models:
15. Command: Costs $0.0015 per 1000 input tokens and $0.0020 per 1000 output tokens.
16. Note: Model customization and Provisioned Throughput are not supported for Cohere models on Amazon Bedrock​​.
17. Stability AI Models:
18. SDXL0.8: Priced per image, with $0.018 per image for standard quality and $0.036 for premium quality at resolutions of 512×512 or smaller. For larger images, the pricing is $0.036 for standard and $0.072 for premium quality.
19. Provisioned Throughput for SDXL1.0 model ranges from $46.18 to $49.86 per hour, depending on the commitment period​​.