The CometAI Platform is UT Dallas’s enterprise generative AI tool, designed to empower the UTD community to innovate across disciplines. Through a flexible, cost-efficient, chat-based interface, CometAI enables users to engage in conversational AI, customize and share responses, and interact with uploaded documents. It's architecture supports vendor independence, allowing seamless selection and switching between multiple AI models, such as those from Anthropic, Google, Meta, Mistral, NVIDIA, OpenAI & xAI without being locked into a single provider.
To access CometAI, open your web browser and visit
https://cometai.utdallas.edu/.
This article covers the various user-facing CometAI API endpoints available to users. It includes a detailed description of each endpoint, example API requests and URLs, sample request bodies, expected response formats, and guidance on how to test and view responses using Postman. The article also explains the expected data returned by each endpoint, common success and error responses, and provides examples of what a successful API call looks like.
This article explains how to access & use the CometAI API Feature. It covers the three types of API keys available in CometAI - Owner Keys, System Keys, and Delegate Keys - including their purpose, when to use each key type, and best practices for selecting the appropriate key based on your use case.
This article guides on how users can choose to store conversations either locally on their device or on the cloud for cross-device access
This article provides information about CometAI model lifecycle status, outlining upcoming deprecations and End-of-Life (EOL) dates to help users plan for continuity. It includes a master deprecation table with removal dates and recommended successor models, plus notes explaining key lifecycle definitions and implications of using deprecated models.
This article covers about the current issues and bugs that are present in CometAI platform
This document provides a detailed comparison of the Large Language Models available within the CometAI platform, including those developed by Anthropic, OpenAI, Mistral, Meta, Gemini, Grok, & NVIDIA. Each model is analyzed based on its design purpose, technical specifications, pricing, and optimal use cases to help users make informed decisions based on their unique needs. Review this document from time to time for updated Model Information in CometAI.
Comet AI includes a built-in Prompt Optimization tool to help users rewrite vague or unclear prompts into clear, effective instructions. This feature is especially helpful for students, faculty, and staff who want better results without needing to be prompt engineering experts.
This article covers all the updates happening in CometAI for the Year 2026.
This article provides a comprehensive overview of the advanced configuration features within CometAI. Understanding and utilizing these settings can significantly enhance the model's performance, efficiency, and relevance for academic research, coursework, and development projects.
Learn how to initiate RAG based conversation in CometAI
Discover the differences of UTD's AI options.
This article explains what Custom Instructions, Prompt Templates & Follow-Up Buttons are, how they work, and how they differ so you know when to use each one.
This article guides on how users can create Assistant group interface - NEW feature launched in CometAI 3.0
This article guide provides step-by-step instructions on how to securely log in & log out to the CometAI platform using your UT Dallas NetID and password.
This article guides on how to personalize their experience by choosing between two visual themes: Light and Dark. This setting can be updated at any time from the platform’s settings menu.
A Prompt Template is a reusable structure that simplifies the creation of AI-generated responses for common tasks. Users will be able to define variable placeholders - such as audience, tone, or topic - that can be quickly customized each time the prompt is used. By using templates, users can streamline repetitive tasks like drafting emails or creating social media posts, reducing the need to rewrite similar prompts and improving workflow efficiency.
Follow-Up Buttons are one-click shortcuts that let you run predefined instructions in a conversation. Instead of retyping common follow-up requests, you can create a button that appears automatically, ready to use.
Learn how to create your own RAG assistant, streamline repetitive processes, ensure consistent outputs, and add tailored functionality to your workspace.
Custom Instructions are a set of guidelines established at the beginning of a conversation with CometAI to guide it's behavior. Establishing elements can determine the tone, style and structure to how CometAI responds to prompts.
This article covers about how to export Chats & Conversations in CometAI Platform.
In addition to exporting single messages, CometAI Platform allows you to download an entire conversation, including your prompts and the Language Models responses, into a single, well-structured PowerPoint slide. This feature is ideal for creating a complete transcript of your interaction for records, sharing, or further analysis as required by your faculty or managers.
This document explains how to organize and manage your conversations in Comet AI. It covers creating folders, moving and sorting conversations, renaming items, searching, deleting, and archiving conversations and folders. Use these features to keep your workspace organized and efficient.
This article covers about how to contact the support team from within the CometAI platform
This article guides how to collaborate by letting you share conversations, prompts, and assistants with other users on the platform.
By adjusting parameters like Temperature and Response Length, you can tailor the output to be more creative or precise, and more concise or detailed, depending on your needs. This guide explains how to access and use these settings to optimize your results.
This article covers the Artifacts feature in CometAI, which allows users to create, manage, and share content like code, documents, and diagrams within a unified workspace. It explains how to enable the feature, use the sidebar tools, preview and edit artifacts, and troubleshoot common issues.
This guide explains how to use the Import Conversations and Export Conversations options in the CometAI Platform. Import Conversations let users upload. json (Conversation) files, and Export Conversations let users download all conversations from their current device and browser in a .json file.
Layered Assistants allow you to build a hierarchy of AI assistants that work together to handle user requests. A Layered Assistant uses a parent assistant to evaluate incoming questions and automatically route them to the most appropriate child assistant for a response.
The Microsoft 365 integration allows CometAI to work with your Microsoft account services - such as OneDrive and SharePoint - directly inside conversations and assistants. Once connected, you can attach files from OneDrive or SharePoint, and ask CometAI to summarize or generate. CometAI only acts when you request it, and it stays within your existing Microsoft 365 permissions.
This document provides a high-level comparison of leading language models currently available within the CometAI platform. The models, developed by Anthropic, OpenAI, Mistral, and Meta, are evaluated across performance strengths, cost-efficiency, and best-fit use cases for academic and enterprise environments. This guide is designed to help faculty and staff select the most appropriate model for tasks such as summarization, writing, planning, coding, and visual document interpretation.
Read this article to become familiar with CometAI features.
The Visual Workflow Builder allows users to create structured, reusable workflows that guide how an AI assistant performs tasks.
The Web Search feature allows CometAI to retrieve and reference information from the internet when generating responses. By combining AI-generated responses with relevant web results, Web Search helps provide more current information, recent updates, and additional sources beyond the model's built-in knowledge.