Web browser
chatgpt.com
Go to chatgpt.com and log in with the SSO option. When logged in, look for "The University of Mar…" in the bottom-left corner.
Verify Your UMD Account
Confirm you are logged in with your UMD account in each tool you use.
Important: If you are not logged in with your UMD account, you will be on the free tier, which is not secure and is severely limited. On the web, try logging out and back in if the session hangs.
VS Code
Codex plugin
Click the gear icon in the top right of the Codex panel. Look for "The University of Maryland" listed under your account name.
Desktop app
Codex for Mac
Click the settings gear in the bottom left. A list appears — your account name should be at the top, with "The University of Maryland" underneath it.
ChatGPT Usage Limit Notifications
Familiarize yourself with these messages so you can quickly identify when a limit has been hit, what it means, and what to do next.
Advanced Models Limit
When a usage limit is reached, ChatGPT displays a banner near the bottom of the interface showing when advanced features will become available again. Affected models appear grayed out and cannot be selected until the limit resets.
Bottom banner: limit reached; fallback model in use
Model dropdown: advanced options grayed out with reset date
Image Generation Limit
When you reach your image generation quota, ChatGPT shows a notification that image creation is temporarily unavailable and gives the reset date and time. The banner may also indicate that advanced models or image-generation features are restricted until the limit period expires.
Note: EDU accounts include a set number of image generations (currently 3). After those are used, generations draw from your available balance.
Image generation quota exhausted — limit resets tomorrow at 10:29 AM
Deep Research Limit
When the Deep Research quota is exhausted, ChatGPT displays a rate limit notification and blocks new Deep Research requests until the limit resets. Each Deep Research task shows a countdown before it starts, as shown below. The number (for example, 52) is the number of seconds until the task begins automatically.
Deep Research countdown before autostart — 52 is the number of seconds until the task begins automatically
Tracking ChatGPT Usage
You can review your ChatGPT usage directly from the signed-in ChatGPT interface.
How to check
chatgpt.com
-
While signed in, click the UMD logo branding
at the bottom-left of the screen.
-
Select Personalization.
- Select Usage.
Usage Limits screen — select Usage and review your monthly usage limit
Work & Codex Usage and Limits
Work is ChatGPT's agentic mode for longer, multi-step tasks and finished deliverables — available alongside regular Chat on the web and in the desktop app. Work and Codex draw from the same monthly credit allowance. Credit consumption is token-based and varies depending on the model, task complexity, input and output size, reasoning effort, speed settings, and overall workload. If you use both Work and Codex, their combined usage counts against the same monthly allowance. Heavy usage in Work can therefore reduce the credits available for Codex, and vice versa. Check your remaining monthly allowance before starting larger Work or Codex tasks.
Codex in VS Code
When your Codex message quota is exhausted, the VS Code extension displays a notification indicating no additional requests can be submitted. You can click View Usage to see where you are in your metering period.
VS Code Codex panel — quota exhausted, request more access from your admin
Codex Efficiency Tips
Reduce unnecessary consumption
6 tips
- Choose an appropriate model. Start with a lower-cost model when it can handle the task, and move to a higher-capability model when the work requires it.
- Set speed to Standard. Fast prioritizes faster execution and may consume credits at a higher rate than Standard. Use Fast when faster turnaround is important for the task.
- Choose the appropriate thinking level. Start with Medium for routine tasks, and use higher reasoning levels when the task benefits from deeper analysis.
- Use
/statusin VS Code. This shows where you are on your usage metering period. To review your full Work & Codex usage history, go to ChatGPT → Settings → Usage — see Tracking Work & Codex Usage below. - Confirm you are logged in. You should see your
@umd.eduaccount in the plugin. If not, you are on the free, unsecured tier. - Check settings after every Codex app update. Updates may silently reset your model, speed, and thinking level back to the highest options.
Tracking Work & Codex Usage
You can review your Work and Codex usage directly in ChatGPT. Use the Usage page to check your current monthly allowance and review where credits are being consumed.
How to check
Settings → Usage
- In ChatGPT, open Settings → Usage. You can also go directly to chatgpt.com/settings/usage.
- On the Overview tab, check your current monthly allowance, credits used, and remaining balance for the current period.
-
On the Analytics tab, review your personal Work and Codex usage history. You can group usage by surface, model, reasoning, or speed to better understand where credits are being consumed.
- You can go directly to chatgpt.com/settings/usage?tab=analytics
- Note: Recent Work and Codex activity may take time to appear in your usage history.
-
Codex in VS Code only: Use
/statusin the VS Code Codex extension to check where you are in your current metering period.
