# Amazon — Optimizing cost and latency with Amazon Bedrock prompt caching

- Company: Amazon (amazon.com)
- Announced: 2026-09-15T16:18:19+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://aws.amazon.com/blogs/machine-learning/optimizing-cost-and-latency-with-amazon-bedrock-prompt-caching/
- Record: https://forck.live/items/10960-optimizing-cost-and-latency-with-amazon-bedrock-prompt-caching
- Subject: Bedrock / Nova
- Pricing: reduced input token costs by up to 90 percent on cache hits

Amazon explains six prompt-caching patterns using the Bedrock Converse API. Cached input tokens can cost up to 90% less on cache hits; the saving does not apply to the entire request or bill.

## Evidence

Verbatim from https://aws.amazon.com/blogs/machine-learning/optimizing-cost-and-latency-with-amazon-bedrock-prompt-caching/:

> Prompt caching in Amazon Bedrock can reduce your input token costs by up to 90 percent when you repeatedly send the same context to foundation models, based on Amazon Bedrock prompt caching pricing.

---

Record: https://forck.live/items/10960-optimizing-cost-and-latency-with-amazon-bedrock-prompt-caching
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md