Home of the Uedu platformShow a desktop notification when the AI TA finishes replying
Notify me when classmates post messages in the forum
Play an alert sound whenever there is a new notification
This is anAI Agent service—you describe the goal, and it reads files, edits files, runs commands, then adjusts its approach based on the results. The model runs on GPUs on campus,so your code never leaves campus..
It is not just for coding—data analysis, document organization, and course learning are all possible. It is powered by a single GPU, so it can only serve a small number of users at the same time—it may queue, but it is completely ours.
Connecting to Uedu Code takes three steps and about 10 minutes total. You do not need any experience with server setup.
Step 1：Create an API key
You need a Uedu account to create an API key—any email address can register, and you do not need a credit card or any company account.
Sign in to create an API keyThe model information, live system status, and usage limits below can be viewed without signing in.
Steps 2 and 3：Install opencode、Generate the config file
The installation commands and config file contents vary by operating system, so they are all placed on the setup guide page; there, only the steps for your chosen installation method will be shown.
View the full guideIn the project folder, run opencode and you are done. Make sure the bottom-right corner shows Uedu Code — if it shows another name, your code is being sent to a service outside campus. Enter /models to switch back.
Please wait a moment.
We believe users have the right to know what they are talking to. Below are the models currently used by this service, including what they do well and what they cannot do.
| Model | Devstral Small 2 Devstral-Small-2-24B-Instruct-2512 |
|---|---|
| Developer | Mistral AI × All Hands AI（法國） |
| Model size | 240 億參數（24B） |
| License | Apache 2.0 |
| Where it runs | GPUs in the campus server room; data stays on campus |
Why we chose it
What it does well
Its limitations
This is the model currently in use, and it may be replaced later. No matter which one it is replaced with, the name you use in the config file will always be uedu-code, so you do not need to change any settings when the model changes.
We believe an edtech platform should make it clear what users leave behind. The notes below are not hidden in the terms.
What gets logged
What these logs are used for
Your rights
Research consent is optional. After you log in, you can sign or withdraw in the consent center. You can still use this service normally without signing—the logs will still be kept for system operations, but they will not be included in research analysis.
Reminder: do not give passwords, API keys, or personal private data to the AI Agent to handle. This is not a special restriction for this service; it is a habit you should have when using any AI tool.