ChatGPT Usage Limits Now Shared With Codex: What Changed

If you use ChatGPT for drafting emails and also lean on Codex for coding help, pay attention: your ChatGPT usage limits just changed. OpenAI merged the two quotas into one shared pool, according to a Hacker News thread that pointed to OpenAI's own help article. That means the tokens you burn writing a blog outline now compete with the tokens you burn generating code. We dug into what changed, who it helps, how to try it in 15 minutes, and the catch you should know before your next heavy writing or coding day.

1. What's new

Until now, ChatGPT and Codex ran on separate usage limits. You could hammer Codex all day without touching your ChatGPT writing quota, and vice versa.

That separation is gone. OpenAI now measures both products against a single shared allowance, as described in ChatGPT and Codex usage limits merging. So the new ChatGPT usage limits apply across both tools at once.

In practice, heavy coding sessions in Codex can eat into the quota you'd normally reserve for writing in ChatGPT. That's a real behavior change, not just fine print.

2. Who this helps

This mostly helps people who only use one of the two tools lightly. If you write in ChatGPT most days and only touch Codex occasionally, your effective limit for writing may feel roughly the same, because you're not drawing from Codex's share.

It can also simplify things for teams that found two separate quotas confusing to track. One shared number is, at least conceptually, easier to reason about than two.

However, if you split your time evenly between writing and coding, you may notice your ChatGPT usage limits feel tighter than before, because both activities now draw from the same bucket.

3. Try it in 15 minutes

First, open ChatGPT and check your current usage status in account settings. Note where you stand before testing anything.

Next, run a normal writing task, for example drafting a short blog outline. Then switch to OpenAI Codex and run a small coding task, like generating a function.

Finally, go back and check your usage numbers again. If both activities moved the same meter, you've just confirmed the merged ChatGPT usage limits in action on your own account.

4. The catch

Here's the catch: OpenAI measures usage in tokens, and tokens aren't something most users can easily see or predict. As one commenter on the Hacker News thread put it, "your usage will be measured in tokens, which we do not define, nor give you a way to measure them."

That same commenter added that "The amount of tokens you consume will depend on things you can't control." So even if you understand the new shared-quota concept, predicting exactly how fast you'll hit your ChatGPT usage limits stays murky.

That said, this isn't unique to OpenAI. Token-based billing is common across AI tools, but merging two previously separate quotas into one makes the uncertainty feel bigger, because now two different workflows compete for the same unclear budget.

5. How it shows up in AI as a Game now

We keep our tool catalog in sync with real changes like this one. The description of ChatGPT in our catalog updated automatically to reflect it: "All-round assistant for text, ideas and drafts; usage limits now shared with Codex."

That line now shows up everywhere ChatGPT appears, on its tool page, across All links, and in every class path or specialist track that references it. So if you're working through our writing tools list, you'll see this context baked right in.

6. What the community is saying

The Hacker News discussion around this change got heated fast. Several commenters focused less on the merge itself and more on how little visibility users get into token consumption, calling the whole token system opaque by design.

Others pointed out that subscription terms can change without much warning, and that annual subscribers are the ones most exposed if usage limits shift mid-contract. It's worth reading the original thread if you want the full range of reactions.

Bottom line

The merged ChatGPT usage limits are a real shift, not a minor tweak. If you only lean on one of ChatGPT or Codex, you probably won't notice much. But if you use both regularly for writing and coding, test your own usage pattern soon, because the shared quota means your habits in one tool now directly affect what's left for the other.