AWS bedrock documentation change
Summary
Updated documentation to clarify Provisioned Throughput concepts, added detailed pricing factors, commitment duration explanations, and restructured workflow steps. Added new sections about MU token processing capabilities and operational procedures.
Security assessment
The changes focus on clarifying billing mechanics, capacity planning, and operational workflows. No security vulnerabilities, access controls, or encryption mechanisms are mentioned. The note about billing continuation until deletion relates to cost management rather than security.
Diff
diff --git a/bedrock/latest/userguide/prov-throughput.md b/bedrock/latest/userguide/prov-throughput.md index 2e8816abd..979124ee7 100644 --- a//bedrock/latest/userguide/prov-throughput.md +++ b//bedrock/latest/userguide/prov-throughput.md @@ -5 +5 @@ -# Provisioned Throughput +# Increase model invocation capacity with Provisioned Throughput in Amazon Bedrock @@ -7 +7 @@ -When you configure Provisioned Throughput for a model, you receive a level of throughput at a fixed cost. +**Throughput** refers to the number and rate of inputs and outputs that a model processes and returns. You can purchase **Provisioned Throughput** to provision a higher level of throughput for a model at a fixed cost. If you customized a model, you must purchase Provisioned Throughput to be able to use it. @@ -9 +9 @@ When you configure Provisioned Throughput for a model, you receive a level of th -You can use Provisioned Throughput with Amazon and third-party base models, and with customized models. +You're billed hourly for a Provisioned Throughput that you purchase. For detailed information about pricing, see [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing). The price per hour depends on the following factors: @@ -11 +11 @@ You can use Provisioned Throughput with Amazon and third-party base models, and -Provisioned Throughput pricing varies depending on the model that you use and the level of commitment you choose. You receive a discounted rate when you commit to a longer period of time. For details about pricing for each model, see the [Model providers](https://console.aws.amazon.com/bedrock/home#/providers) page in the Amazon Bedrock console. + 1. The model that you choose (for custom models, pricing is the same as the base model that it was customized from). @@ -13 +13,5 @@ Provisioned Throughput pricing varies depending on the model that you use and th -Your options for throughput for a model differ depending on whether you run inference on a base model or a custom model. + 2. The number of Model Units (MUs) that you specify for the Provisioned Throughput. An MU delivers a specific throughput level for the specified model. The throughput level of an MU specifies the following: + + * The number of input tokens that an MU can process across all requests within a span of one minute. + + * The number of output tokens that an MU can generate across all requests within a span of one minute. @@ -17 +21,3 @@ Your options for throughput for a model differ depending on whether you run infe -In the AWS GovCloud (US) region, you can only purchase Provisioned Throughput for custom models with no commitment. +For more information about what an MU specifies, contact your AWS account manager. + + 3. The duration of time you commit to keeping the Provisioned Throughput. The longer the commitment duration, the more discounted the hourly price becomes. You can choose between the following levels of commitment: @@ -19,5 +25 @@ In the AWS GovCloud (US) region, you can only purchase Provisioned Throughput fo -Pricing option | Base model | Custom model ----|---|--- -Provisioned Throughput, no commitment (hourly pricing) | Not available | Available (maximum 2 Provisioned Throughputs per account) -Provisioned Throughput, 1 month commitment | Available | Available -Provisioned Throughput, 6 month commitment | Available | Available + * No commitment – You can delete the Provisioned Throughput at any time. @@ -25 +27 @@ Provisioned Throughput, 6 month commitment | Available | Available -You specify Provisioned Throughput in Model Units (MU). A model unit delivers a specific throughput level for the specified model. The throughput level of a MU for a given Text model specifies the following: + * 1 month – You can't delete the Provisioned Throughput until the one month commitment term is over. @@ -27 +29,3 @@ You specify Provisioned Throughput in Model Units (MU). A model unit delivers a - * **The total number of input tokens per minute** – The number of input tokens that an MU can process across all requests within a span of one minute. + * 6 months – You can't delete the Provisioned Throughput until the six month commitment term is over. + +###### Note @@ -29 +33 @@ You specify Provisioned Throughput in Model Units (MU). A model unit delivers a - * **The total number of output tokens per minute** – The number of output tokens that an MU can generate across all requests within a span of one minute. +Billing continues until you delete the Provisioned Throughput. @@ -34 +38 @@ You specify Provisioned Throughput in Model Units (MU). A model unit delivers a -Model unit quotas depend on the level of commitment you specify for the Provisioned Throughput. +The following steps outline the process of setting up and using Provisioned Throughput. @@ -36 +40 @@ Model unit quotas depend on the level of commitment you specify for the Provisio - * For custom models with no commitment, a quota of one model unit is available for each Provisioned Throughput. You can create up to two Provisioned Throughputs per account. + 1. Determine the number of MUs you wish to purchase for a Provisioned Throughput and the amount of time for which you want to commit to using the Provisioned Throughput. @@ -38 +42,3 @@ Model unit quotas depend on the level of commitment you specify for the Provisio - * For base or custom models with commitment, there is a default quota of 0 model units. To request an increase, use the [limit increase form](https://console.aws.amazon.com/support/home#/case/create?issueType=service-limit-increase). + 2. Purchase Provisioned Throughput for a base or custom model. + + 3. After the provisioned model is created, you can use it to [run model inference](./inference.html). @@ -45 +51,7 @@ Model unit quotas depend on the level of commitment you specify for the Provisio - * [Supported Region and models for Provisioned Throughput](./pt-supported.html) + * [Supported Region and models for Provisioned Throughput](./prov-thru-supported.html) + + * [Prerequisites for Provisioned Throughput](./prov-thru-prereq.html) + + * [Purchase a Provisioned Throughput for an Amazon Bedrock model](./prov-thru-purchase.html) + + * [View information about a Provisioned Throughput](./prov-thru-info.html) @@ -47 +59 @@ Model unit quotas depend on the level of commitment you specify for the Provisio - * [Procedures](./prov-thru-workflow.html) + * [Modify a Provisioned Throughput](./prov-thru-edit.html) @@ -49 +61 @@ Model unit quotas depend on the level of commitment you specify for the Provisio - * [Permissions](./prov-thru-permissions.html) + * [Use a Provisioned Throughput with an Amazon Bedrock resource](./prov-thru-use.html) @@ -51 +63 @@ Model unit quotas depend on the level of commitment you specify for the Provisio - * [Provisioned Throughput console procedures](./prov-cap-console.html) + * [Delete a Provisioned Throughput or cancel auto renew](./prov-thru-delete.html) @@ -53 +65 @@ Model unit quotas depend on the level of commitment you specify for the Provisio - * [Using the Provisioned Throughput API](./prov-thru-api.html) + * [Code examples for Provisioned Throughput](./prov-thru-code-examples.html)