OpenAI extends shared prompt caching to 12 hours
AI summary
A Reddit post citing AI Economics Weekly reports that OpenAI now lets shared prompt context remain cached for up to 12 hours, up from 30 minutes. The report says the longer cache is intended to reduce input costs for slow workflows, but the claim has not been independently verified in the supplied material.
In-depth insight
The reported extension directly changes how long developers can reuse shared prompt context in slow API workflows, making it a notable OpenAI cost-control update.
This source is a digest of several news items (“7 OpenAI updates from this week that cut your A…”). Find this story in the original.