prompt_cache_retention becomes prompt_cache_options.ttl on GPT-6

Q: How do I fix this when moving to gpt-6*?

Replace prompt_cache_retention with prompt_cache_options.ttl. Check the allowed TTL values in the docs.

Behavior change Applies to gpt-6* on the OpenAI Chat Completions and OpenAI Responses API.

OpenAI's docs describe this behavior change but don't quote the exact message, so we don't either.

How to fix it

Replace prompt_cache_retention with prompt_cache_options.ttl. Check the allowed TTL values in the docs.

Before

{
  "messages": [
    {
      "content": "Summarize this ticket.",
      "role": "user"
    }
  ],
  "model": "gpt-4o-2024-08-06",
  "prompt_cache_retention": "24h"
}

After (GPT-6 Luna)

{
  "messages": [
    {
      "content": "Summarize this ticket.",
      "role": "user"
    }
  ],
  "model": "gpt-6-luna",
  "prompt_cache_options": {
    "ttl": "24h"
  }
}

Check your whole request in the linter