By the Smili Phone team
If you use both ChatGPT for everyday questions and Codex for coding help, your two tools are about to draw from the same pool. OpenAI says that starting in December, ChatGPT and Codex will share one usage allowance on Plus, Pro, and Business plans. The change comes as OpenAI folds its Work features into the main ChatGPT experience, so chat and coding assistance live under one roof, and one budget.
This is not a shutdown of either tool. It is a change in how paid usage is counted, and it matters most if you use both heavily in the same period.
What is actually changing
Today, many paid users think of ChatGPT and Codex as separate allowances. From December, OpenAI's updated Help Center article says the allowance will be shared across both. Usage is measured in tokens, which is the unit AI models use to count text and code. Advanced models and complex tasks use more tokens, so a long coding session can eat through the pool faster than a few quick questions.
You will see the combined picture in Settings, then Usage, with remaining usage and reset times in one place. That stops a common surprise: thinking you have plenty left in one tool while the other has used the shared budget. OpenAI expects most users will not wait for resets more often than now, but heavy users running long coding jobs and long chats in the same week are most likely to feel the difference.
What stays the same
Instant chat does not draw from the usage budget, according to OpenAI. That is useful to know, because quick questions are not the thing most likely to drain your allowance. The heavy lifting is long documents, deep research style answers, and complex coding tasks that ask the model to read and write a lot of code.
OpenAI also recently doubled usage limits on higher priced plans ahead of the merger, according to coverage of the announcement. Anthropic already combined its Chat and Cowork features in Claude, so the direction is clear: fewer separate apps, and one account covering chatting, working, and coding.
How to make one allowance last
- Start light. Use Instant chat or a smaller model for quick questions, and save advanced models for work that really needs them.
- Be specific in coding prompts. A clear task and the exact outcome you want usually take fewer rounds than a vague request followed by fixes.
- Break big jobs into smaller steps. Ask for one feature at a time, check it, then move on.
- Check Settings, then Usage, before a big session. If you are close to the limit, do the important work first.
- If you hit the limit, OpenAI says you can wait for the reset, upgrade, or buy credits. For most people, waiting is the sensible first choice.
What this means for everyday users
For a casual user, this change is mostly housekeeping. You chat, you occasionally ask for code help, and the shared pool is plenty. For a student, a freelancer, or a small business owner who leans on both tools every day, it is worth a little planning. Think of it like a shared data plan: two devices, one allowance, and the heavy user sets the pace.
If you use AI tools for work or study in Escondido and are unsure which plan fits your routine, bring your questions by the shop. We can check the settings on your phone or laptop with you.
Related: OpenAI's agent push at DevDay · Google's agent gets its own email
