Try a group top to bottom. The first question fills the cache; the rephrasings underneath should hit it.
Refunds
Hours
Shipping
Savings dashboard
live totals across every query this cache has servedTotal requests
2
Cache hit rate
50%
LLM calls avoided
1
Total saved
$0.000042
Time saved
2.2 s
Cache contents
similarity threshold 0.44| Stored query | Hits | Original cost |
|---|---|---|
What is your refund policy? gpt-4o-mini | 1 | $0.000042 |