ChatGPT appears limited to Instant across all models after Codex/work quota is exhausted; Pro reports a usage limit

What issue are you seeing?

After my remaining Codex/work quota reached 0, I started experiencing restrictions in the ChatGPT chat interface on the same account:

• All models appear to provide only Instant-level behavior, even when I select a different model or mode.
• The Pro option is blocked by a message indicating that I have reached a usage limit.

The impact extends beyond Codex: I am also unable to use the expected capabilities in regular ChatGPT conversations.

I observed this after my Codex/work quota was exhausted. I cannot confirm whether the two are causally related, but the timing suggests a possible issue with how usage limits are applied across Codex and ChatGPT.

What steps can reproduce the bug?

The sequence that led to the issue on my account was:

1. Use Codex/work until the remaining quota reaches 0.
2. Open the regular ChatGPT chat interface on the same account.
3. Select different models or modes and send messages.
4. Observe that responses appear limited to Instant-level behavior.
5. Select Pro and see a message indicating that a usage limit has been reached.

What is the expected behavior?

I would expect exhausting my Codex/work quota to restrict Codex usage without affecting access to ChatGPT models or modes, unless a shared limit explicitly applies.

If this is expected behavior, please clarify which limit is shared and why exhausting Codex/work usage also restricts Pro in ChatGPT.

Additional information

Could you investigate whether a Codex/work usage restriction is being incorrectly applied to ChatGPT model access on my account?

No source file, test, or entry point is named; start by reproducing the quota sequence across Codex and the ChatGPT model selector on the same account. Done means determining whether the limits are shared and documenting or correcting the behavior so Codex exhaustion does not unexpectedly restrict ChatGPT access.