OpenAI API adds 12-hour prompt caching for persistent agents
jeppa35 · reddit · 2026-10-21
OpenAI's developer account announced 12-hour prompt caching in the API, up from the 30-minute default. You can mix both in a single request: 12h for shared instructions, 30 minutes for newer conversation content. Persistent agents that reuse context all day — like dots — can keep shared instructions, tools, and context cached longer and keep paying the lower cached-input rate even when requests are hours apart. The accompanying Reddit post adds the cost logic: when the same context is reused repeatedly across a day, savings from cache hits significantly outweigh the extra 12-hour caching cost.
Related event: OpenAI API Adds 12-Hour Prompt Caching for Long-Running Agents(4 posts)→
More from Models
- Ask your AI agent its favorite color — it apparently has one — every · 2026-10-21
- Google demos Gemini 4 Argon building a 3D periodic table synced to Tom Lehrer's 'The Elements' — sundarpichai · 2026-10-21
- MiniMax follow-up: M3.1 Flash teased after GTC release wave — MaziyarPanahi · 2026-10-21
- Claude Max 5x limits feel generous: heavy user can't burn through weekly quota — Angaisb_ · 2026-10-21
- General Intuition announces API coming soon, pitching itself as the frontier lab for acting in space and time — chris_j_paxton · 2026-10-21
- Stronger Model, Weaker Product: Dev Defaults to Fable 6.1 Sol Ultra Over 5.1 — curious_vii · 2026-10-21