What are ChatGPT Plus usage limits and when do they reset?

Last updated: 26 September 2026 · Deutsch

OpenAI names no fixed message count for ChatGPT Plus. Its help pages say only that limits depend on your plan and model, may include message caps when demand is high, and change over time. There is no reset time that applies to everyone: when you hit a limit, ChatGPT tells you and shows when that allowance resets, if it knows. Until then you can switch to another model or wait.

This page sticks to what OpenAI’s own help and pricing pages say on the date above. OpenAI changes models and limits often, so check again before you rely on a detail.

What OpenAI says per plan

Free. Unlimited everyday text chats with GPT-5.6 Luna, within safeguards against abuse. File uploads, image generation, voice, data analysis and other tools have their own limits. OpenAI gives no numbers for them.
Go. The same unlimited everyday text chats, plus more messages with tools, more uploads and more images than Free. Again no numbers. For harder questions, Go has Think, which also runs on GPT-5.6 Luna.
Plus. $20 a month. Higher limits than Free and the Medium and High thinking levels of GPT-5.6 Sol; Extra High and Pro are not included. The pricing page links “usage limits apply” to OpenAI’s model article. That article says Thinking limits depend on your plan, but names no figure for Plus.
Pro. Two tiers: Pro $100 with five times the Plus usage, Pro $200 with twenty times. Here OpenAI does give numbers for the Pro models in chat. On Pro $200, GPT-6 Pro allows 200 messages per week; GPT-5.6 Sol Pro has a separate 170 per day, and both together are capped at 200 per day. On Pro $100, both share one allowance of 50 messages per week. Since 10 September 2026, OpenAI has paused new sign-ups and upgrades to Pro $200.

Why you find other numbers online

OpenAI names no fixed number for Plus. Its help pages say Plus has higher limits than Free, may include message caps during high demand, and that these limits vary with system conditions.

Many sites list message counts per three hours or per week for Plus. Earlier versions of OpenAI’s help pages did name caps like that for older models. The current pages no longer do so for Plus chat. Treat any such figure as out of date unless OpenAI’s current page states it.

The pages themselves move, too: the model article changed its title and its web address with the switch to GPT-5.6. What we don’t know is whether OpenAI still uses fixed caps for Plus behind the scenes; it does not publish them.

When limits reset

There is no reset time that applies to everyone. When you reach a limit, ChatGPT shows when that allowance resets, if it has that information. OpenAI Support cannot reset your limits.

OpenAI’s current pages describe no fixed window length for Plus chat. What they do say:

Each allowance has its own reset. The Pro-model allowances above run per day or per week; other plans can differ.
ChatGPT shows when an allowance resets, when that information is available. You see it in the app, not on a public page.
On Free, a tool limit pauses that tool until it resets, and ChatGPT shows the timing in the product.
If you hit a limit on Free and upgrade to Plus, Pro or Business, your usage resets.
Support cannot reset Codex limits either. If access has not returned after the stated reset time, or you think your use was counted wrong, Support will look into it.

How you notice a limit

ChatGPT tells you when you reach a limit. On Pro, a model whose allowance is used up may be unavailable until its reset. Neither means your account is restricted.

A restriction is something else: OpenAI’s safeguards against misuse can now and then block use for a while, and ChatGPT tells you when that happens. If you think it was a mistake, contact Support.

What happens at the limit

If you reach a Thinking limit, ChatGPT may carry on with another available model. You can also pick another model yourself or wait until the reset time it shows.

On Plus and Pro, Instant does not switch up to a higher thinking level by itself, and asking it to “think harder” does not either. Pick the level in the model menu.
On Pro $200, when the weekly GPT-6 Pro allowance runs out, ChatGPT switches to GPT-5.6 Thinking at Medium. You can still choose GPT-5.6 Sol Pro while its daily allowance lasts.
On Pro $100 and on Business, switching between the two Pro models gives you no extra messages: they share one allowance.

There is no setting to raise or skip a model’s allowance.

Codex has its own allowance

OpenAI states that the chat limits above apply to chat only. ChatGPT Work and Codex share an allowance of their own, apart from chat, with its own usage and credit rules. Plus includes GPT-6 Astra in Work and Codex, though not in chat.

So a long Codex session does not use up your chat messages, and the other way round. How to check the Codex allowance, with its five-hour and weekly windows, is in How do I check my Codex usage and remaining limit?

ChatGPT plan or API

A ChatGPT plan and the OpenAI API are two different products. OpenAI says plainly that API use is not part of Plus and is billed on its own. Your plan pays for ChatGPT; the API bills your developer account for what you use.

Do Codex and the API count against your chat limits? No. Codex and ChatGPT Work have their own allowance, separate from chat. API use is not part of ChatGPT Plus at all and is billed on its own.

How to track API spend is covered in AI API costs at a glance and How much is the OpenAI API costing me this month?.

In AI Cockpit

The ChatGPT/Codex card sits in the Mac menu bar, in the app on iPhone and iPad, and, passed on by the iPhone app, on Apple Watch. You sign in with your ChatGPT account inside the app.

Your plan name as a label on the card, for example “Plus”.
The Codex allowance: the five-hour and the weekly window, each with the share used and its reset time.
A warning colour by default at 75 and 90 per cent on the Mac, at 80 and 95 on iPhone and Apple Watch, for those windows.
Your Codex credit balance.
If you enter what you pay for the plan each month, the app counts it in the cost total.

What it does not show: how many chat messages you have left, chat Thinking limits or the Pro-model allowances. The data the card reads from OpenAI covers the Codex allowance only, and OpenAI publishes no chat figures the app could read.

More on the Codex windows is in Codex quotas and reset times. Short answers to other questions are on the FAQ page.