# \`reasoning.encrypted\_content\` support for Chat Completions API?

**URL:** <https://community.openai.com/t/reasoning-encrypted-content-support-for-chat-completions-api/1287547>\
**Category:** Feedback\
**Tags:** reasoning\
**Created:** [June 13, 2025, 5:17pm UTC](https://community.openai.com/t/reasoning-encrypted-content-support-for-chat-completions-api/1287547 "2025-06-13T17:17:45Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![manouchehri](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/manouchehri/32/194066_2.png) [@manouchehri](https://community.openai.com/u/manouchehri)\
**Post date:** [June 13, 2025, 5:17pm UTC](https://community.openai.com/t/reasoning-encrypted-content-support-for-chat-completions-api/1287547/1 "2025-06-13T17:17:45Z")

</div>

The lack of encrypted content support in the Chat Completions API is a huge headache for some of us. The Response API is great for some situations, but it’s not universally better.

I don’t understand why it can’t be added; when you use `reasoning.encrypted_content` , the request becomes stateless, meaning the Responses API is a bit pointless.

Related to [`n` support for Responses API?](https://community.openai.com/t/n-support-for-responses-api/1280343).

---

<div class="post-metadata">

**Author:** ![\_j](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/_j/32/766292_2.png) [@\_j](https://community.openai.com/u/_j)\
**Post date:** [June 13, 2025, 7:26pm UTC](https://community.openai.com/t/reasoning-encrypted-content-support-for-chat-completions-api/1287547/2 "2025-06-13T19:26:55Z")

</div>

Chat Completions doesn’t have an out-of-band method of passing a conversation history with different items or events. It would take a convolution of the format along the lines of new role messages, and new parameters to “include” something like summaries and reasoning that would break all existing code.

Almost like a new API endpoint. Voila!

But one that is not a step back in features and a huge code burden that yet treats the developer as an untrusted party that is a consumer of tool products that need nanny instructions that cannot be controlled. And features that need biometric scans and government IDs in order to get “dangerous” streaming messages.

---

<div class="post-metadata">

**Author:** ![manouchehri](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/manouchehri/32/194066_2.png) [@manouchehri](https://community.openai.com/u/manouchehri)\
**Post date:** [June 13, 2025, 7:47pm UTC](https://community.openai.com/t/reasoning-encrypted-content-support-for-chat-completions-api/1287547/3 "2025-06-13T19:47:04Z")

</div>

`reasoning.encrypted_content` disables storing the conversation history. There is _no_ “out-of-band” method used with the Responses API when you use `reasoning.encrypted_content`; that is the whole point of it.

OpenAI _has_ literally added multiple new role types to messages before. e.g. `tool` and `developer` did not used to exist.

Any client that cannot handle the `reasoning` response fields could just… _not_ add the optional `"include": ["reasoning.encrypted_content"]` field to their Chat Completion API request. If _you_ include an optional parameter that breaks your code, then you should probably ask `gpt-4.1-nano-2025-04-14` for help.