How to check your ChatGPT usage on a Mac (all four ways)

To check your ChatGPT usage on a Mac, open ChatGPT and inspect the model picker for a reset time, or read the notice shown when you reach a limit. ChatGPT does not give consumer accounts one complete usage dashboard, so browser tools and menu bar trackers help when you want a clearer view before the limit arrives.

The short answer

ChatGPT exposes usage in several places rather than one universal counter:

1. The model picker can show when a model-specific allowance resets.
2. Limit notices explain which model or feature has reached its cap.
3. Browser tools can collect those signals into a more convenient display.
4. A macOS menu bar tracker keeps the allowance and reset time visible outside ChatGPT.

The right one depends on whether you want to check your usage or simply know it. The first two answer a question you have to remember to ask. The last two put the answer where you will see it before you need to ask.

First, understand what ChatGPT is counting

ChatGPT does not have one account-wide bucket that every action draws from. Models and tools can have separate allowances, different reset windows, and different fallback behavior. Text models, image generation, file uploads, deep research, and voice can each be governed differently.

OpenAI's model guidance for ChatGPT describes limits by plan and model, while the Free Tier FAQ explains that tools can have limits separate from the main text allowance. A reset shown for one model is not necessarily the reset for every ChatGPT feature.

Limits also change as models, plans, and demand change. The time shown inside your own account is more useful than a fixed number copied from an old article. This guide focuses on where to find live information instead of promising a message count that may already be stale.

That fragmented design is why people often think they are missing a settings page. They are not. Consumer ChatGPT accounts currently surface the relevant information around the model or tool affected, not in one complete account dashboard.

Way 1: check the model picker

Open ChatGPT in a browser or the Mac app, start or open a conversation, then select the model name near the top. When ChatGPT exposes a reset time for a model-specific allowance, this is one of the places it appears.

The model picker is useful before a demanding task. It can tell you whether the model you want is available and whether another model is the sensible fallback. It does not combine every model and feature into one total.

If no reset time is displayed, do not assume the model is unlimited. It may share a broader plan allowance, or ChatGPT may only show the relevant information when you approach the cap.

Best for: checking a particular model before starting work. Worst for: understanding the whole account at a glance.

Way 2: read the in-product limit notice

When you reach a cap, ChatGPT shows a notice in the conversation. The wording varies, but it normally identifies the affected model or tool, explains what remains available, and gives a reset time when OpenAI provides one.

This is the most authoritative answer for the limit you just hit because it reflects the account, plan, and feature in use. Its drawback is timing: it tells you at the wall, not while you still have room to plan around it.

Best for: a definitive explanation after a cap is reached. Worst for: avoiding the interruption.

Way 3: browser based trackers

Browser extensions can turn ChatGPT's scattered usage signals into a badge, popup, or small dashboard. Some record reset timestamps as they appear; others read account data exposed to the signed-in web session.

This is convenient if nearly all your ChatGPT work happens in a browser. But it normally depends on the browser being open, and it may stop refreshing when no ChatGPT tab is active.

Check permissions before installing one. A tool that can read pages on chatgpt.com may also be able to see conversation content. Prefer narrow permissions, local processing, and a clear privacy policy.

Best for: browser-first workflows. Worst for: people who mainly use the Mac app or want usage visible everywhere.

Way 4: the macOS menu bar

The fourth option is to stop checking altogether and put the number somewhere you cannot help but see it. On a Mac that means the menu bar, which is visible from every app, at all times, whatever you happen to be doing.

This is how tokn.watch works. Connect your ChatGPT account once and the current usage appears in the menu bar. If you also use Claude, both providers can sit side by side even though their limits reset on different schedules.

The ChatGPT connection guide explains setup. Credentials stay on the Mac, and the app connects directly to the provider rather than sending the session through a tokn.watch server.

Best for: knowing where you stand without asking. Worst for: people who genuinely never want another icon in their menu bar, which is a fair preference.

Which method should you use?

If you only need to know whether a particular model is available, the model picker is enough. If limits regularly interrupt work, an always-visible tracker answers the more useful question: how close am I before I start?

What to do when you are close to the limit

Seeing the number is only useful if it changes what you do. Three things actually help:

• Switch models before you run out, not after. Lighter models draw far less from your allowance. Using the most capable model for routine work is the most common way people burn a session early.
• Separate model and tool limits. If image generation is capped, ordinary chat may still work. If one model is capped, another may remain available.
• Use the displayed reset time. ChatGPT windows do not all reset at midnight. See when ChatGPT usage resets.

If limits keep interrupting work, compare the cost of those interruptions with the next plan. Our guides ask whether ChatGPT Plus is worth it and whether ChatGPT Pro is worth it without assuming an upgrade is always the answer.

Frequently asked questions