Is this actually better value for getting things done? Instead of throwing another $20 at credits that seem to get gobbled up in seconds, just switch to a second ChatGPT Plus account on a laptop when there’s a backlog of things you want to do. When one account hits its weekly limit, switch to the other, then cancel the second subscription once you’re caught up.
I can relate to this concern, and I think many users encounter the same question once they begin using Codex more heavily.
Before spending too much on additional credits, it may be worth reviewing how Codex is being used across your projects. A practical approach is to estimate your typical monthly workload first, then choose the plan or usage option that best matches that level of activity.
It can also help to review some of the guidance around using Codex more efficiently. For example, you can ask Codex itself to help you prepare or refine your project’s AGENTS.md so that instructions are clearer, repeated context is reduced, and unnecessary work is avoided. Small improvements in project structure and instructions can make a noticeable difference over time.
There are several ways to balance plans, credits, and workflow efficiency, so I would probably optimize the workflow first before adding another account.
Hopefully this helps a little. And it would still be useful to have clearer guidance from the OpenAI team on the most efficient approach for users with occasional periods of heavy Codex usage.