GPT 5.5 Instant on API .. just via chat-latest? Pricing?

bsommerhalder: > is it just the same cost (but presumed to use fewer tokens) as regular 5.5 The only way that there could be fewer tokens is if the final answer had less text. There is no billed internal reasoning to be “smarter” or more efficient. What is not the same cost is the same input and same output: