--- title: Better prompt caching for GPT‑6 | daily.dev description: OpenAI has shipped an improved prompt caching system for the GPT-6 model family, offering higher default cache hit rates with discounts up to 90% on cached... canonical: https://daily.dev/posts/better-prompt-caching-for-gpt-6-jkbbgeye0 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Better prompt caching for GPT‑6 | daily.dev og:description: OpenAI has shipped an improved prompt caching system for the GPT-6 model family, offering higher default cache hit rates with discounts up to 90% on cached... og:url: https://daily.dev/posts/better-prompt-caching-for-gpt-6-jkbbgeye0 og:image: https://api.daily.dev/og/posts/jKbBgeYE0.png og:image:alt: Better prompt caching for GPT‑6 og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Better prompt caching for GPT‑6 **[OpenAI](https://daily.dev/sources/openai)** · 4 min read · 0 upvotes · 0 comments ## Summary OpenAI has shipped an improved prompt caching system for the GPT-6 model family, offering higher default cache hit rates with discounts up to 90% on cached input tokens for shared prefixes reused within a 30-minute window. New tools include a Prompt Caching Dashboard for tracking hit rates and input composition, a diagnostics tool that identifies why a cache miss occurred (e.g., changed tools or settings), explicit cache breakpoints to control which prefixes get cached, the ability to change reasoning effort mid-conversation without breaking cache, guidance for managing tool/instruction changes via append-only updates, and cache prewarming to reduce latency by preparing shared context ahead of user requests. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How does prompt caching work with GPT-6 and how much discount do I get on cached tokens? GPT-6 caches shared context such as instructions, tool definitions, and prior turns to avoid recomputation, giving discounts of up to 90% on cached input tokens. Cache discounts now apply to eligible shared prefixes reused within a 30-minute window, with higher hit rates by default compared to earlier caching behavior. _daily.dev surfaces updates like this for developers optimizing agent latency and API costs._ ### How can I diagnose why my OpenAI API request had a prompt cache miss? Use the prompt caching diagnostics tool to compare a request against a recent response and identify what changed, such as the model, tools, settings, or input, that prevented cache reuse. The response reports a reason code like tools_changed along with the estimated number of tokens affected, helping size the impact and guide optimization. _Developers debugging latency spikes can track tooling like this via daily.dev._ ### How do I change reasoning effort in GPT-6 without losing my prompt cache? On GPT-6 models, reasoning effort can be changed between responses without breaking cache by appending a configuration_update rather than changing the request-level reasoning effort. This lets an agent raise effort for a harder task or lower it for a routine follow-up while preserving reusable cached context. _daily.dev helps engineers tuning agent reasoning and cost tradeoffs stay current._ ## Similar posts on daily.dev - [Prompt Caching Explained](https://daily.dev/posts/prompt-caching-explained-mhomvre8o) · DigitalOcean Community · 2 upvotes · 0 comments - [Monitor prompt caching to optimize your token usage](https://daily.dev/posts/monitor-prompt-caching-to-optimize-your-token-usage-bwzwycrgn) · Datadog · 0 upvotes · 0 comments - [Prompt Caching for Anthropic and OpenAI Models: Building Cost-Efficient AI Systems](https://daily.dev/posts/prompt-caching-for-anthropic-and-openai-models-building-cost-efficient-ai-systems-mjzh92xkq) · DigitalOcean · 2 upvotes · 0 comments - [Prompt Caching Techniques](https://daily.dev/posts/prompt-caching-techniques-rlnexmcrg) · PromptLayer Blog · 1 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#openai](https://daily.dev/tags/openai), [#gpt](https://daily.dev/tags/gpt), [#ai-inference](https://daily.dev/tags/ai-inference) [View this post on daily.dev](https://daily.dev/posts/better-prompt-caching-for-gpt-6-jkbbgeye0) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Better prompt caching for GPT‑6","url":"https://daily.dev/posts/better-prompt-caching-for-gpt-6-jkbbgeye0","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/better-prompt-caching-for-gpt-6-jkbbgeye0"},"datePublished":"2026-09-22T20:15:15.935Z","dateModified":"2026-09-23T16:07:48.992Z","description":"OpenAI has shipped an improved prompt caching system for the GPT-6 model family, offering higher default cache hit rates with discounts up to 90% on cached...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/cabe69b261c04edca5bbc3ce26b42aae?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/cabe69b261c04edca5bbc3ce26b42aae?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"OpenAI","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"OpenAI","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/984b3632316f4afaa3d5213d5c64fe65","url":"https://daily.dev/sources/openai"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/better-prompt-caching-for-gpt-6-jkbbgeye0","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,openai,gpt,ai-inference","timeRequired":"PT4M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"OpenAI","item":"https://daily.dev/sources/openai"},{"@type":"ListItem","position":3,"name":"Better prompt caching for GPT‑6"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/better-prompt-caching-for-gpt-6-jkbbgeye0#faq","mainEntity":[{"@type":"Question","name":"How does prompt caching work with GPT-6 and how much discount do I get on cached tokens?","acceptedAnswer":{"@type":"Answer","text":"GPT-6 caches shared context such as instructions, tool definitions, and prior turns to avoid recomputation, giving discounts of up to 90% on cached input tokens. Cache discounts now apply to eligible shared prefixes reused within a 30-minute window, with higher hit rates by default compared to earlier caching behavior. daily.dev surfaces updates like this for developers optimizing agent latency and API costs."}},{"@type":"Question","name":"How can I diagnose why my OpenAI API request had a prompt cache miss?","acceptedAnswer":{"@type":"Answer","text":"Use the prompt caching diagnostics tool to compare a request against a recent response and identify what changed, such as the model, tools, settings, or input, that prevented cache reuse. The response reports a reason code like tools_changed along with the estimated number of tokens affected, helping size the impact and guide optimization. Developers debugging latency spikes can track tooling like this via daily.dev."}},{"@type":"Question","name":"How do I change reasoning effort in GPT-6 without losing my prompt cache?","acceptedAnswer":{"@type":"Answer","text":"On GPT-6 models, reasoning effort can be changed between responses without breaking cache by appending a configuration_update rather than changing the request-level reasoning effort. This lets an agent raise effort for a harder task or lower it for a routine follow-up while preserving reusable cached context. daily.dev helps engineers tuning agent reasoning and cost tradeoffs stay current."}}]} ```