OpenAI expands caching in Codex, Work and API
OpenAI expands caching in Codex, Work and API
OpenAI made caching persistent in Codex and Work and added a 12-hour cache retention option in the API at twice the cost of 30-minute caching.
OpenAI expands caching in Codex, Work and API
OpenAI has improved caching behavior across its products. In Codex and Work, cache is now retained persistently, so a session no longer consumes a large amount of limits on refilling the cache when a user steps away for a few hours.
In the API, OpenAI introduced an option to retain cache for 12 hours. This extended retention costs twice as much as writing to the 30-minute cache.
Why it matters
AnalysisPersistent caching reduces wasted limits during long sessions and gives developers more control over context retention in the API. For teams using AI tools daily, this can make workflows cheaper and more stable.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
Codex Skills for Beginners: Automating AI Agent Workflows
A comprehensive hands-on guide to building and executing Codex Skills: SKILL.md anatomy, explicit vs. auto-triggering, skills vs. scripts vs. tools, Skills API, and safety boundaries.
ChatGPT Updates: Plugin Development and Codex Workflows
Comprehensive engineering guide to creating, packaging, testing, and submitting plugins for ChatGPT and Codex: manifest specification, MCP servers, lifecycle hooks, custom marketplaces, and OpenAI review.
OpenAI GPT (Flagship Models of the GPT Series)
The primary universal line of large language models from OpenAI (GPT-4, GPT-4o). Optimized for complex text analysis, programming, creativity, and daily intellectual tasks.
DeepSeek releases DeepInfer, an open-source inference framework
DeepSeek has published DeepInfer, an Apache-2.0 inference framework used in its production services, with optimized kernels, MoE parallelism, speculative decoding, deterministic mode, and structured outputs.
Stanford Releases SpaceX Lecture on How Grok Bot Works
The 45-minute session covers persistent AI agents, agent harness design, and context management for long-running tasks.
Frontier Models
The most powerful class of artificial intelligence at the forefront of global research (Claude 3.7 Sonnet, OpenAI o3/GPT-4.5, Gemini 2.0 Pro), defining the limits of modern reasoning, autonomy, and coding capabilities.