GPT-6 Sol and Luna may keep prompt cache when reasoning effort changes mid-chat

Dan McAteer @daniel_mac8Originally posted 2026.09.22

Open original on X

For GPT-6 Astra, changing the reasoning effort mid-conversation does NOT invalidate the prompt cache. Too many AI engineers are unaware of this. OpenAI also implies in their docs that it will be the case for forthcoming GPT-6 model families like Sol and Luna. Take advantage.

This post quotes another post: x.com/daniel_mac8/status/2102140412647321802