Bedrock Knowledgebase RetrieveAndGenerate requests per second quota

Hi Bhaskar,

Please log a support case and provide as much details as possible which can explain why this is required and can't be worked around.

Support engineering team would need to work with service/engineering team and should be able to bump up the limit once they understand the requirement completely. In my case, we were able to get this quota increased by requesting through support case explaining the requirement.

Let me know how it goes.

Feel free to reach out if you have additional questions, happy to help.

-- Abhishek

The quota for Knowledge Bases RetrieveAndGenerate requests per second in Amazon Bedrock is indeed set at a default value and is marked as "Not Adjustable" in the service quotas documentation. This means that unlike some other Bedrock quotas, this particular limit cannot be increased through the standard Service Quotas console process.

If you're experiencing limitations due to this quota and need an increase, you'll need to contact AWS Support directly. This cannot be done through the regular Service Quotas interface. When contacting AWS Support:

1. Create a support case through the AWS Support Center
2. Specify the exact quota you need increased (Knowledge Bases RetrieveAndGenerate requests per second)
3. Include details about your use case and why you need a higher limit
4. Mention the AWS Region for which you need the increase

Be aware that due to high demand for Bedrock services, AWS typically prioritizes quota increase requests for customers who are actively consuming their existing quota allocation. Your request might be denied if you're not currently utilizing a significant portion of your existing quota.

While waiting for a response from AWS Support, you might consider implementing request throttling or adding small delays between requests in your application to manage the flow of requests to the Bedrock API and prevent hitting the rate limits.
Sources
Amazon Bedrock endpoints and quotas - AWS General Reference
Question regarding Error invoking RetrieveAndGenerateCommand: Your request rate is too high. Reduce the frequency of request: | AWS re:Post
Request an increase for Amazon Bedrock quotas - Amazon Bedrock
Can not increate invokeModel request to Claude 3.5 Sonnet v2 | AWS re:Post