Skip to main content

Using Open WebUI

Overview​

Open WebUI is ASU Research Computing's deployment of the open-source, feature-rich, and user-friendly AI platform. It provides a ChatGPT-like interface for interacting with Large Language Models hosted by Research Computing.

Open WebUI supports OpenAI-compatible protocols, making it a powerful, provider-agnostic AI interface that works seamlessly with the models available through Research Computing.

Accessing Open WebUI​

Visit ai.rc.asu.edu and log in with your ASURITE credentials.

Key Features​

Chat Interface​

  • Conversational AI: Interact with LLMs through a familiar chat interface similar to ChatGPT.
  • Multiple Models: Switch between different available models based on your needs.
  • Conversation History: Your chat history is saved for future reference.

Document & Data Interaction​

  • File Uploads: Upload documents and have the AI analyze or summarize them.
  • Code Assistance: Get help with coding tasks, debugging, and code explanation.
  • Research Support: Use AI to help with literature review, writing, and analysis.

Customization​

  • System Prompts: Customize how the AI responds by setting system prompts.
  • Model Parameters: Adjust temperature and other parameters for different use cases.

Sub-agents​

Open WebUI supports sub-agents: the model can hand a self-contained piece of work to a background helper that runs its own tool-driven conversation and reports back only its conclusion. This keeps a long research thread readable instead of filling your main chat with intermediate tool output.

  • Enabling it: sub-agents must be switched on globally by an administrator, and enabled per model under Workspace → Models → Edit. They also require native function calling in the UI.
  • Using it: ask for work that splits naturally — "research these four approaches and compare them" — and a capable model delegates rather than doing everything in one thread.
  • Limits: a sub-agent cannot delegate again, so work never fans out without bound. There are administrator-set caps on how many run concurrently and how long each may iterate.
Open WebUI cannot edit your files

Even with sub-agents, this is a chat interface — it has no access to your project. For an assistant that reads and writes files, see Getting Started with AI. Open WebUI is excellent for the thinking that happens before you open an editor: shaping a problem, explaining an error, or drafting the prompt you will paste into your coding assistant.

Available Models​

The models available through Open WebUI are managed by Research Computing. To see the current list of available models:

  1. Log in to ai.rc.asu.edu
  2. Click on the model selector dropdown in the chat interface
  3. Browse the available models

You can also view available models in the Voyager portal under the LLM Access tab.

Tips for Effective Use​

  • Be Specific: Provide clear, detailed prompts for better responses.
  • Use Context: Include relevant background information in your questions.
  • Iterate: Refine your prompts based on the responses you receive.
  • Try Different Models: Different models may perform better for different tasks.

Learn More​

For detailed documentation on Open WebUI features, visit the Open WebUI Documentation.