AWS general documentation change
Summary
Added new service quotas for Amazon Nova 2 Lite model customization and Anthropic Claude 3 Haiku token limits.
Security assessment
Changes define resource limits and deployment parameters without addressing vulnerabilities or security controls. No security implications as these are standard service quota updates.
Diff
diff --git a/general/latest/gr/bedrock.md b/general/latest/gr/bedrock.md index ba2f17591..4e4b0615b 100644 --- a//general/latest/gr/bedrock.md +++ b//general/latest/gr/bedrock.md @@ -344,0 +345 @@ Name | Default | Adjustable | Description +(Model customization) Sum of on demand custom model deployment requests per minute for Amazon Nova 2 Lite | Each supported Region: 2,000 | No | The sum of input and output on-demand custom model deployment requests per minute submitted to the Converse, ConverseStream, InvokeModel, and InvokeModelWithResponseStream actions for Amazon Nova 2 Lite @@ -348,0 +350 @@ Name | Default | Adjustable | Description +(Model customization) Sum of on demand custom model deployment tokens per day for Amazon Nova 2 Lite | Each supported Region: 5,760,000,000 | No | The sum of input and output on-demand custom model deployment tokens per day submitted to the Converse, ConverseStream, InvokeModel, and InvokeModelWithResponseStream actions for Amazon Nova 2 Lite @@ -352,0 +355 @@ Name | Default | Adjustable | Description +(Model customization) Sum of on demand custom model deployment tokens per minute for Amazon Nova 2 Lite | Each supported Region: 4,000,000 | No | The sum of input and output on-demand custom model deployment tokens per minute submitted to the Converse, ConverseStream, InvokeModel, and InvokeModelWithResponseStream actions for Amazon Nova 2 Lite @@ -356,0 +360 @@ Name | Default | Adjustable | Description +(Model customization) Sum of training and validation records for a Amazon Nova 2 Lite Fine-tuning job | Each supported Region: 20,000 | [Yes](https://console.aws.amazon.com/servicequotas/home/services/bedrock/quotas/L-82EE9228) | The maximum combined number of training and validation records allowed for a Amazon Nova 2 Lite Fine-tuning job. @@ -710,0 +715 @@ Model invocation max tokens per day for Amazon Nova Premier V1 (doubled for cros +Model invocation max tokens per day for Anthropic Claude 3 Haiku (doubled for cross-region calls) | Each supported Region: 432,000,000 | No | Daily maximum tokens for model inference for Anthropic Claude 3 Haiku. Combines sum of input and output tokens across all requests to Converse, ConverseStream, InvokeModel and InvokeModelWithResponseStream. Doubled for cross-region calls; not applicable in case of approved TPM increase.