> ## Documentation Index
> Fetch the complete documentation index at: https://docs.weflayr.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Prompt caching analyzer
> Providers do not always perfectly follow what your caching settings asked for, so your LLM calls may cost more than what you expect. This page tracks those misses.
Every figure compares your actual caching, exactly as it ran, against what those same settings should have done. PS: none of the improvements the [Optimisation engine](/product/cost-optimisation) recommends are applied here.
The *Open caching analyzer* button is at the bottom of the Prompt caching card:
## Call details
Click a row to open the call's own page.
### Cache lineage
It displays the calls behind the cache entry this call read: the call that **wrote** the cache entry, the calls that read it since, and this call last. PS: Some providers' caching have their Time-To-Live (TTL) refreshed on every hit. In this case the penultimate call is labeled *Cache last refreshed by*.
### Prompt
Visualize which parts of the prompt have been correctly cached and which have been missed.
