OpenAI's 12-Hour Prompt Cache: Break-Even Math for GPT-6 Astra, Luna and Sol
OpenAI's new 12-hour prompt cache costs 2.5x input to write. It beats re-writing the 30-minute cache after two lapses on Astra, Luna and Sol. The math.
· 7 minTopic
Prompt Caching
Coverage tagged prompt caching.
- Articles
- 2
- Desk
- 1
Prompt Caching Explained: Cutting Agent Costs Without Touching the Model
Your agent re-sends the same long prefix on every step and pays to recompute it. Here is the KV mechanism, the ordering rule it forces, and the break-even math.
· 11 min