A reported change to the OpenAI Codex codebase appears to add broader support for a plan identified internally as promax. The change involves Codex’s backend, the server-side systems and logic that handle accounts, limits, and plan data—but it does not show that the plan is publicly available.
In a post on X, @TokenGremlin reports that ProMax was added across the Codex backend and that its display name is now mapped to “Pro (Max).” The attached repository-diff screenshot shows a change titled “Add Pro Max plan support and update Pro display names,” along with a new ProMax enum entry.
Image credit: @TokenGremlin on X
What the reported Codex changes show
The report describes promax being recognized in several parts of Codex’s backend. These include authentication, account responses, backend rate limits, generated schemas, client types, experimental context eligibility, and usage and credit handling.
In practical terms, these are the kinds of systems that allow software to identify a plan, return it in account data, apply plan-specific limits, represent it in shared data definitions and client code, and account for it in eligibility or usage logic. Taken together, the reported additions suggest that ProMax is being wired into more of Codex’s internal plan-handling plumbing rather than existing only as an isolated identifier.
The screenshot also describes Pro Max as being included in paid-plan tooltips and usage-limit guidance for purchasing credits. Those details indicate that related user-facing text or guidance was considered in the change, but they do not establish that the labels or guidance are currently visible to every Codex user.
How the reported plan names are mapped
The displayed names in the reported change differ from the internal identifiers:
proliteis displayed as “Pro”prois displayed as “Pro (More)”promaxis displayed as “Pro (Max)”
The change therefore appears to revise the labels for more than just the newly reported ProMax entry. In the screenshot, the earlier mapping of pro to “Pro” is replaced with “Pro (More),” while prolite becomes “Pro.” The new promax value is mapped to “Pro (Max).”
These labels describe how the plans are represented in the reported code change. They do not, by themselves, explain differences in pricing, quotas, model access, performance, or eligibility.
Why backend support does not confirm public availability
Backend recognition is not the same as a public product launch. A plan can be represented in authentication, account schemas, rate-limit logic, and client types before users can select it or before a rollout is complete.
The supplied evidence does not confirm any of the following:
that users can currently choose ProMax in Codex;
that OpenAI has publicly announced the plan;
what ProMax would cost;
which quotas or rate limits would apply;
which models or features it would include;
who would be eligible in practice; or
when, or whether, a broader rollout will occur.
The reported changes make ProMax look more fully integrated into the systems shown in the repository diff. They do not establish public availability. For now, the most supported description is that ProMax appears to be receiving broader backend support and a defined “Pro (Max)” display label, while the user-facing launch status remains unconfirmed.
A related ChatGPT Pro report examines possible higher-tier code references without confirming a public launch.
0 comments
No approved comments yet. You can start the conversation.
Leave a comment