agentVIDEO

OpenAI improves prompt caching in the GPT-6 API

GPT-6 now has higher cache-hit rates by default, so more input tokens get cached-input discounts of up to 90%, helping agents run faster and cost less.

OpenAI Developers

@OpenAIDevsArchived 2026.09.23Originally posted 2026.09.23

OpenAI Developers @OpenAIDevsOriginally posted 2026.09.23

Open original on X

We’ve improved prompt caching in the API for GPT-6, helping agents run faster and cost less. Higher cache-hit rates by default mean more input tokens benefit from cached-input discounts of up to 90%. https://t.co/NZse8QKr8G

x.com/openaidevs/status/2102506476401258678Archived 2026.09.23 10:42