This page describes how RWTHgpt is structured and what functions the service offers.
Detailed information can be found in the linked articles.

RWTHgpt works similarly to other AI chatbots as a text-based dialogue system. As a user, you can ask the AI questions, assign text-based tasks, or, depending on the AI tool or language model, have multimodal content (text & images) analyzed.
AI content and responses may be incorrect, so please always check important information.
Tips on how to best submit questions and tasks to the AI are covered in our article Ask Better, Answer Better: Prompt Engineering with RWTHgpt.

The KI:connect web interface, in which RWTHgpt is embedded, offers various functions and information, which are explained in detail in the corresponding articles.

• Legal information
• Language model selection
• Chat functions
• API keys and access

In the article Legal Information, you can learn more about the legal background of RWTHgpt, the privacy policy and terms of use, as well as information on legally compliant use. Information on this is also linked directly in the web interface.

In the article Language Model Selection, you can learn how to switch between the available language models and what limitations there are. Newly added language models and other innovations are published on our IT Center Changes Blog.

In the article Chat Functions, you can learn about the additional functions enabled by the chat and how you can use them. These include editing sent prompts, chat export and editing options, document and image upload options, and analyses.

In the article API Keys and Access, you can learn how to generate an API key and how to select a model from those offered.