- Newest
- Most votes
- Most comments
Hello,
You can check billing information under Bedrock Service and Bedrock:
- you can also verify the details on Cost explorer by Selecting - Dimension type - Usage type , Service - Bedrock
The pricing for AWS Bedrock agents is not explicitly detailed in the provided information. However, based on the available details, here's what we can understand about the pricing structure:
-
When using AWS Bedrock agents, you are primarily charged for the inference calls made to the foundation models. This is based on the volume of input and output tokens processed.
-
The InvokeAgent API itself is not charged separately. This means that the act of calling the agent doesn't incur a direct cost.
-
If you've configured your agent with a knowledge base, there may be additional charges related to the storage and retrieval of information from that knowledge base.
-
The $0.04 charge you mentioned for the Bedrock service could be related to other aspects of using the service, such as storage of your agent configuration, processing of the agent's logic, or other underlying infrastructure costs.
-
If you're using provisioned throughput for your models, that would also factor into the pricing.
Unfortunately, without more specific information from AWS about the breakdown of charges for Bedrock agents, it's difficult to pinpoint exactly what the $0.04 charge represents. It's possible that this is a service fee for using the agent functionality, but this is not explicitly stated in the available information.
To get a more detailed understanding of your charges, you may want to:
- Check the detailed billing information in your AWS account.
- Contact AWS Support for a breakdown of the charges.
- Look for any updates to the Amazon Bedrock pricing documentation, as new information may become available as the service evolves.
Remember that AWS services often have various components to their pricing, and some aspects may not be immediately apparent in the general pricing overview.
Sources
Agents for Amazon Bedrock is now available with improved control of orchestration and visibility into reasoning | AWS News Blog
Amazon Bedrock pricing - Amazon Bedrock
answered 2 years ago
There are two majors factors in determining the cost of Bedrock agents:
- algorithm (Amazon Nova, Claude, Llama, etc.)
- mode of invocation
Algorithm Different algorithms have different pricing and will deliver different results. In our experience the it's worth comparing the following 4 algorithms gives a good performance cross section:
- Amazon Nova
- Anthropic's Claude
- Meta's Llama
- DeepSeek
Invocation mode There are two primary modes of invocation: on-demand and batch. By batching invocation requests you can access reduced pricing. Additionally cache reads and writes can be lower cost.
We've created a blog post with a table showcasing the pricing difference between different models and modes: https://www.metaltoad.com/blog/aws-bedrock-pricing
The following is a screenshot and little bit of an awkward copy and paste with the data:
Anthropic models Price per 1M input tokens Price per 1M output tokens Price per 1M input tokens (batch) Price per 1M output tokens (batch) Price per 1M input tokens (cache read) Price per 1M input tokens (cache write) Claude 3.7 Sonnet $3.00 $15.00 N/A N/A $0.30 $3.75 Claude 3.5 Sonnet $3.00 $15.00 $1.50 $7.50 $0.30 $3.75 Claude 3.5 Haiku $0.80 $4.00 $0.50 $2.50 $1.00 $0.08 Amazon Nova models Price per 1M input tokens Price per 1M output tokens Price per 1M input tokens (batch) Price per 1M output tokens (batch) Price per 1M input tokens (cache read) Amazon Nova Micro $0.04 $0.14 $0.02 $0.07 $0.01 - Amazon Nova Lite $0.06 $0.24 $0.03 $0.12 $0.02 - Amazon Nova Pro $0.80 $3.20 $0.40 $1.60 $0.20 - Amazon Nova Pro (w/ latency optimized inference) $1.00 $4.00 N/A N/A N/A - Amazon Nova Premier $2.50 $12.50 $1.25 $6.25 N/A - Meta models Price per 1M input tokens Price per 1M output tokens Price per 1M input tokens (batch) Price per 1M output tokens (batch) NA NA Llama 4 Maverick 17B $0.24 $0.97 $0.12 $0.49 - - Llama 4 Scout 17B $0.17 $0.66 $0.09 $0.33 - - Llama 3.3 Instruct (70B) $0.72 $0.72 $0.36 $0.36 - - DeepSeek Models Price per 1M input tokens Price per 1M output tokens NA NA NA NA DeepSeek-R1 $1.35 $5.40 - - - -
answered a year ago
Relevant content
asked 2 years ago
asked 2 years ago
- AWS OFFICIALUpdated 6 months ago
- AWS OFFICIALUpdated 6 months ago
- AWS OFFICIALUpdated a month ago
