Use case
GPT-5.4 Mini prompt caching
prompt caching notes for gpt-5.4-mini: the rate card, a worked volume, and when to send this ID. A sample of 1 million input tokens plus 200 thousand output tokens is about $0.330 on corouter (input $0.150, output $0.180).
Cache-read rate
Cache read for gpt-5.4-mini lists at $0.075 and is $0.015 on corouter per million tokens. Uncached input is $0.150.
Only the input that actually hits cache is billed at the cache-read rate. This page does not promise your prompt will hit.
Where this ID sits
GPT-5.4 Mini cuts the list rate and the window to 400K, and it has no vision tag. Send gpt-5.4-mini for shorter reasoning jobs.
Questions
- Which model ID should the request send?
- Send gpt-5.4-mini. Do not put a different name in the model field.
- What is the minimum top-up?
- Prepaid credits, $5 minimum top-up, no subscription tier.