Use case
GPT-5.4 Nano prompt caching
prompt caching notes for gpt-5.4-nano: 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 $0.090 on corouter (input $0.040, output $0.050).
Cache-read rate
Cache read for gpt-5.4-nano lists at $0.020 and is $0.0040 on corouter per million tokens. Uncached input is $0.040.
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 Nano is the lowest GPT-5.4 list rate, with a 400K window and vision. Send gpt-5.4-nano for high-volume chat and image passes.
Questions
- Which model ID should the request send?
- Send gpt-5.4-nano. 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.