{
  "bedrock_facts": [
    {
      "id": "brk_provisioned_throughput",
      "model_id": null,
      "model_family": "all models",
      "category": "provisioned throughput",
      "description": "Bedrock Provisioned Throughput lets customers purchase dedicated model capacity measured in Model Units (MUs) — guaranteed tokens-processed-per-minute (input+output) — billed hourly regardless of usage, as an alternative to per-request on-demand pricing, with 1-month or 6-month commitment terms.",
      "source_url": "https://repost.aws/questions/QUFY3RP2J8RJqbt4u6QyKIdw/can-somebody-explain-bedrock-provisioned-throughput",
      "created_at": "2026-07-02 06:03:26",
      "cite_as": "https://subagentbedrock.com/api/bedrock-facts/brk_provisioned_throughput"
    },
    {
      "id": "brk_provisioned_when",
      "model_id": null,
      "model_family": "all models",
      "category": "provisioned throughput",
      "description": "Provisioned Throughput is cost-effective for high-volume, consistent workloads where guaranteed capacity without throttling is required and on-demand costs would exceed the provisioned hourly rate; on-demand remains preferable for spiky or low/unpredictable volume.",
      "source_url": "https://inventivehq.com/blog/aws-bedrock-pricing-guide",
      "created_at": "2026-07-02 06:03:26",
      "cite_as": "https://subagentbedrock.com/api/bedrock-facts/brk_provisioned_when"
    }
  ]
}