OpenAI Merges ChatGPT and Codex Limits, Subscribers Cry Foul

One gauge for two very different tools

Starting soon, Plus, Pro, and Business subscribers will see a single usage meter covering both ChatGPT and Codex. According to OpenAI's updated documentation, more complex tasks will consume more of that allowance, while ordinary Instant conversations remain outside it.

The logic behind the merge is straightforward: OpenAI wants ChatGPT to be a single platform for everything from quick questions to full-blown coding projects. Bringing ChatGPT Work's autonomous capabilities into the main chat experience means those projects draw from the same pool as everything else.

That convenience comes with a trade-off. A subscriber who uses ChatGPT for research and occasionally fires up Codex for a side project will now see both activities competing for the same resources. For heavy Codex users, the shared limit could feel like a downgrade even if OpenAI insists otherwise.

OpenAI says most subscribers shouldn't find themselves waiting for limits to reset more often than they do today. But the company hasn't disclosed the size of the new allowances, which makes that reassurance difficult to evaluate. It's essentially asking customers to trust that the subscription they already pay for won't become less useful.

Why subscribers are worried

The frustration isn't really about the mechanics of the change. It's about predictability. People build workflows around ChatGPT and Codex, and mixing the two meters makes it harder to plan usage.

For most people, ChatGPT is still a straightforward assistant. They aren't thinking about how many tokens a software development agent burns while building an app. The new shared allowance forces that awareness on them, whether they want it or not.

The Instant chat carve-out is an important protection. It means a quick question won't suddenly become impossible because someone got carried away with Codex. But the more complex features that make ChatGPT Work appealing are exactly the ones that will eat into the shared limit.

OpenAI is clearly betting that the convenience of one platform outweighs the annoyance of one meter. That bet might pay off for power users who want everything in one place. For everyday subscribers, it's less obvious.

What OpenAI hasn't said

The biggest unknown is the size of the new limits. Without that number, subscribers can't tell whether they're gaining flexibility or losing headroom.

OpenAI's documentation update confirms the merge but leaves the allowance sizes unspecified. That's a significant omission. A shared limit that's generous enough could be a non-issue. A tight one could push heavy users toward higher tiers or competing tools.

The company has also not detailed how different task types will be weighted. A complex Codex session might consume far more than a long ChatGPT conversation, and without clear guidance, users will have to learn the new rules by hitting them.

For now, the change is a trust exercise. OpenAI is asking paying customers to accept that the subscription they know won't become less valuable once the rules change. Whether that trust holds will depend on the numbers it eventually publishes.

What it means for you

The change applies to ChatGPT Plus, Pro, and Business subscribers in the US, UK, EU, and other markets where these plans are available. OpenAI hasn't announced a specific rollout date or regional exceptions.

If you're on a Plus plan and rarely touch Codex, you may not notice much difference—Instant chats stay outside the shared limit. If you use Codex regularly, keep an eye on your usage once the new meter goes live, because the old separate allowances will no longer apply.

The merge makes sense for OpenAI's platform ambitions, but it puts the burden of proof on the company. Subscribers deserve to know exactly what they're getting before the new meter replaces the old ones. Until OpenAI publishes the numbers, the smart move is to watch your usage closely and plan for less headroom, not more.