Codex for the Classroom

Current Grant Submission Status: The current Codex Classroom Grant Program submission form is now closed. Additional opportunities to apply for Codex for Classroom grants are expected in future iterations of the program as funding and capacity allow.

OpenAI’s Codex is an AI coding tool that can help users write, understand, review, and debug code. U-M Faculty can request Codex for the Classroom, enabling access to Codex for themselves and their students. There are costs associated with this service and a U-M Shortcode is required.

How to Get Started

U-M Faculty can apply for the ITS grant or request Codex for the Classroom via U-M GPT Toolkit.

  1. Log into U-M GPT Toolkit.
  2. Click + Create Key and select Provision Codex for Classroom.
  3. Complete and submit the form that displays. If you are applying for the ITS grant, you'll also need to enter a description of how you will use Codex in your class as part of the form. Your description should address:
    1. How Codex will support course activities, assignments, or learning objectives
    2. Why Codex is appropriate for this use
    3. The approximate number of students expected to participate
    4. Whether you'd be open to being contacted by ITS about your experience

Faculty covering Codex costs through their own Shortcode without a grant do not need to provide this description.

An AI Services team member will contact you with additional information to help you get started.

How Codex for the Classroom Works

Before you begin, it’s important to understand some key information about how U-M’s Codex for the Classroom works and the overall process for setting it up.

  • The Canvas course roster is used to provision your students with Codex API keys. Enrollments sync twice per day from Canvas to account for roster changes.
  • U-M AI Services provides access to Codex models through U-M GPT Toolkit, allowing you to use Codex without managing a separate OpenAI API account. 
  • Codex runs locally on your computer while model requests are sent through the Toolkit gateway.
  • Currently, U-M only supports Codex used from your terminal through the Codex command-line interface (CLI). 

Set Up Codex

Once you’ve been approved for Codex for the Classroom, there are two main steps to set it up.

1. Retrieve your Codex API key

Faculty and students can follow these instructions to retrieve your unique API key.

Keep your API key private. Treat it like a password — do not share it with anyone. Anyone with your key can use your stipend.

2. Set up Codex on your local computer

Faculty and students can follow the setup guide for your operating system:

Best Practices for AI-Assisted Coding

  • Use GPT-5.6-Terra for most of your queries to make the best use of your quota.
  • Review generated code before running it, especially anything that modifies files or installs packages.
  • Keep prompts specific. Clear instructions with relevant context (file names, error messages, expected behavior) produce better results than vague requests.
  • Commit or save your work before letting Codex make changes, so you can easily roll back if needed.
  • Treat AI-generated code the same as code from any other source: test it, and don't assume it's correct just because it runs.

Support & Documentation

Need help? Contact the ITS AI Services team.

For Faculty & Students

For Faculty

For Students