A company has created a custom model by fine-tuning an existing large language model (LLM) from Amazon Bedrock. The company wants to deploy the model to production and use the model to handle a steady rate of requests each minute. Which solution meets these requirements MOST cost-effectively?
Select an option, then click Submit answer.
Reference / correct answer:
Purchase Provisioned Throughput for the model on Amazon Bedrock.
Most accepted answer: D. Purchase Provisioned Throughput for the model on Amazon Bedrock.
Community votes: D=2
Selected Answer: D D. For custom (fine-tuned) models on Amazon Bedrock, deploying to production requires purchasing Provisioned Throughput; on-demand mode is not available for most custom models. Provisioned Throughput reserves dedicated model units and provides a guaranteed, predictable capacity for a steady (not bursty) workload, with discounted pricing over pay-as-you-go options. Option B (on-demand throughput) is best for unpredictable or low-volume workloads, as it charges per token and may be unavailable for custom fine-tuned models, which commonly require Provisioned Throughput for production deployment upvoted 5 times
Selected Answer: D Amazon Bedrock Provisioned Throughput is designed for predictable, sustained workloads. It provides dedicated model capacity and is generally more cost-effective than on-demand throughput when request volume is steady and known in advance. Suitable for production workloads with predictable demands. upvoted 1 times