Yesterday I installed CODEX in Visual Studio to build the app for a leisure activity (personal use). It works very well as commanded, and I am very happy with the outputs compared to Copilot. But I have a concern with weekly limits. There are two places to check, and both have different value which one should I consider? I can not understand.
Welcome to the forum!
You can discuss Codex-related issues here, but the official place to report them is the OpenAI Codex GitHub Issues.
I used ChatGPT to search for similar issues and found several related issues.
It’s worth reviewing existing issues to see if any align with your situation. If none do, consider opening a new report with clear reproduction steps. If you do find a match, adding a 
reaction can help increase its visibility to the developers.
For context, I only currently subscribe to PLUS.
I’ve been using Codex heavily for several weeks now.
The 5-hour limit and the weekly limit are connected because both get used when you use Codex.
Usually, the 5-hour limit runs out first. After that, you need to wait until it renews, usually around 5 hours later.
The weekly limit takes longer to run out, but if you finish that too, then waiting 5 hours will not renew the weekly limit. You need to wait until the weekly limit resets on the stated date.
For the 258k context token limit, that means how much information Codex can keep in mind at one time. If your project becomes too large, especially if you rely on Codex to write code, debug code, read many files, and fix many things, it may start missing or forgetting small details.
So you may need to remind Codex clearly, for example:
“Do not remove this feature.”
“Remember to keep this design.”
“Check this file too.”
“Make sure the previous fix still works.”
Simple way to explain it:
The usage limit is like Codex’s “battery.”
The context token limit is like Codex’s “memory space.”
Even if Codex still has battery, if the memory space gets too full, it may forget small details.
Currently, I’m using GPT-5.5 at medium, and yes, it does run out quite quickly. But just use it. There’s nothing to be scared of. If you run out, just wait until it renews again.
Have fun building.
Welcome to the dev community @Harshal03
The top snippet in you screenshot is what shows the rate limits.
The bottom snippet in your screenshot shows the part of the context window that the current conversation has consumed.