GPT-6 improves prompt caching

KEY POINTS

1. GPT-6 introduces improved prompt caching with higher cache hit rates, new diagnostics, explicit breakpoints, and user controls intended to reduce latency and costs.
2. These features target prompt-heavy workflows that can benefit from more efficient reuse of prompt context.
3. Improved prompt caching can materially lower latency and operating costs for applications that repeatedly reuse prompts, making production deployments more efficient.