Use case
GPT-5.4 prompt caching
prompt caching notes for gpt-5.4: 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 $1.10 on corouter (input $0.500, output $0.600).
Cache-read rate
Cache read for gpt-5.4 lists at $0.250 and is $0.050 on corouter per million tokens. Uncached input is $0.500.
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 is the general 1.05M GPT row beneath the Pro rate and above Mini. Send gpt-5.4 for chat, reasoning, and vision without the Pro multiplier.
Questions
- Which model ID should the request send?
- Send gpt-5.4. 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.