Feature Request: Optional Extra High Reasoning in Existing Regular Chats for Plus Users
Plan: ChatGPT Plus
Product: Regular Chat — not Work or Codex
I’d like to request an option to use Extra High reasoning selectively within an existing Regular Chat conversation.
The problem
GPT-5.6 Sol’s High reasoning is sufficient for most of my work. However, I occasionally need deeper reasoning for multi-file analysis, complex architecture, difficult debugging, and detailed consistency reviews.
I would like to use Extra High for those individual messages without having to move my work to another conversation, Work thread, or Codex session.
Maintaining conversation continuity is particularly important for long-running projects.
Requested behavior
Please allow Plus users to:
-
Select Extra High for an individual difficult message within an existing Regular Chat.
-
Return to High for subsequent ordinary messages.
-
Continue in the same conversation without starting a new thread.
Extra High should be optional, not a replacement for High.
Where technically and commercially feasible, a deeper Max option could also be considered in the future. However, Extra High is my primary request and would already be a meaningful improvement.
Usage and plan considerations
The Usage interface states that the displayed limits are shared across Codex, Work, Workspace Agents, and ChatGPT for Excel, while Chat conversations are not included in that particular shared pool.
I understand that this does not mean Regular Chat is unlimited or that Plus users are entitled to additional compute.
I am not requesting unlimited access, Pro-only models, Work/Codex tools, or any guaranteed usage allowance.
I’m requesting the capability, not prescribing how OpenAI should meter, limit, price, or technically implement it.
OpenAI can determine the appropriate access conditions, safeguards, capacity controls, and plan differentiation.
Why this would help
The main benefit is the ability to match reasoning effort to each task without unnecessarily disrupting an ongoing conversation.
High would remain my normal setting. Extra High would be used only when a particularly difficult message warrants it.
Question for the community:
Would other Plus users find optional per-message Extra High reasoning useful in Regular Chat, particularly for long-running conversations that alternate between routine and complex tasks?
I’m interested in whether others experience the same need for deeper reasoning without switching products or starting a new conversation.