OpenAI is making GPT-6 prompt caching more reliable and controllable, adding higher hit…
This is a AI post classified by Jev as AI infra & evals (a launch), kept by the AI Radar because it carries real work, not commentary.
OpenAI is making GPT-6 prompt caching more reliable and controllable, adding higher hit rates, diagnostics, breakpoints, prewarming, and cache-preserving reasoning changes. Long-running agents often resend instructions, tool definitions, and earlier context, so caching avoids recomputing those prefixes across successive API calls. That reuse can cut cached-input token costs by up to 90%, while a new dashboard exposes cache-hit rates and cached versus uncached token volume. ofcourse, a high cache hit rate is still remains partly an application-design problem. because GPT-6’s improved cachin
Posted by Rohan Paul (157.9k followers) 1 h ago · 8 likes · 1.2k views · view the original post on X. Kept by the AI Radar as AI infra & evals.
More AI work like this
- New on Dedicated Model Inference: canary rollouts. — @togethercompute
- No coincidence — @lennysan
- we are so back? — @demian_ai
- Ornn's B200 Rental Index has reached an all time high. — @OrnnExchange
- Deploying DiffusionGemma-Jev (djev) just got a lot easier. You can now spin up a Jev… — @googlegemma
- Cyber attackers move fast, so defenders have to move faster 🚨 — @cerebras
- Yesterday marked one month anniversary of @DarkbloomAI moving to paid tier on OpenRouter. — @gajesh
- Actually wait, answer quality might have gone up. But 25% savings is definitely holding. — @mmastrac
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 43.6k posts from 5k X accounts over the last 14 days, 1.8k tools, 19 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-23 00:23 UTC. Full method.