Noob qns: How does "memory" work with Codex via VScode extension?

Hi, been using my ChatGPT Plus via chat for coding for like 2 years now.

With the great reviews I saw for Codex CLI and their recently released VSCode extension and allowing connection with the same ChatGPT Plus account, looking to try it out.

But how does the “memory” work between the 3 scenarios (that I can think of) ?:

  1. Project folders in my ChatGPT Plus chat (with existing chats / context that I prefer to continue working off)
  2. Codex CLI via VScode using my ChatGPT Plus
  3. Codex CLI via VScode using API token pricing

Does scenario 2 and 3 able to access the memory / context I have in my chat project folders?

If not, how does it usually work in regards to preserving context?

I read around and understand that prompting will eat up my ChatGPT Plus weekly limits pretty fast and I want to actively manage that since I’m using it alot for work. Refreshing context for scenario 2 and 3 will quickly hit that limit / pay for API tokens.

please advise!

3 Likes

Hi!

Thanks for answering, this is weird, I can’t see your reply on OpenAI’s forums so replying via email.

Appreciate your help, was asking ChatGPT about this same question and it said it’s possible for 2 to share memory with web chat interface LOL.

what is the answer? I’m curious, too.