Prompts
Instructions, context, and techniques for getting useful work from language models.
Articles
2 PUBLISHEDUnderstanding Gemini Context Caching: When Should You Create a Cache?
Gemini can turn repeated context into a named resource. Create a versioned cache, choose its TTL, reuse it across questions, measure cached tokens, and clean it up.
Understanding OpenAI Prompt Caching: Why Does Prefix Order Matter?
OpenAI can automatically reuse eligible prompt prefixes. Put stable content first, keep changing input last, and verify reuse through cached-token telemetry.