LibGuides: AI Tools for Research and Writing: Home

Top Generative AI Chatbots

Anthropic Claude
Claude can process large amounts of information, brainstorm ideas, generate text and code, help learn about subjects, roleplay through situations, simplify text. It requires an email address and login and there's a subscription option ($20/month) for more advanced tools. Claude 3.5 Haiku was trained on data up until July 2024.. The free version lets you have longer conversations with the bot than the others I've listed. Anthropic posts regular updates on their model at https://www.anthropic.com/news
ChatGPT
The original and most recognized chatbot from Open AI. Subscribers ($20/month) get access to the latest model. As of 2025: Free tier users have access to a large range of capabilities with GPT-4o, including access to a series of tools and access to GPTs in theGPT store. Free tier users can now: Search the web and obtain up-to-date information, analyze and extract insights from your data, upload images or files in your prompts, discover and use GPTs, create images in ChatGPT. Due to the computing costs of long conversations with the bot, free users are limited to a certain number of questions overall and within a conversation with the bot.
Copilot
Copilot for Microsoft 365 uses a combination of LLMs, a type of artificial intelligence (AI) algorithm that uses deep learning techniques and vast data sets to understand, summarize, predict, and generate content. These LLMs include pre-trained models, such as Generative Pre-Trained Transformers (GPT) like GPT-4 and GPT-4o, designed to excel in these tasks. Free users have access to GPT-4o. Requires account creation to save conversations. Microsoft will release operating systems and new laptop models with the chatbot built in to their proprietary apps.
Duck.ai
*RECOMMENDED* DuckDuckGo's interface allows you to chat with some of the top AI models. Their goal is to provide private, useful, and optional AI features that don’t track how you use them, store your prompts, or train on your data.
Perplexity
Perplexity was created specifically to answer questions with real-time web information. Allows you to switch between Claude's and OpenAI's most current large language models. Subscribers get access to more ($20/month or $200/year). You must create an account to get access to a limited number of free advanced searches. Due to the computation cost, free users are limited in the amount of questions they can ask per hour and per day.

There is an open source ranking tool for large language models (LLMs) behind the chatbots. The Chatbot Arena Leaderboard is built and maintained by research group LMSYS and the University of California, Berkeley’s Sky Computing Lab. The Chatbot Arena will allow you to prompt two models at once to easily compare their responses. Another leaderboard ranks LLMs based on how much electric power they use per query.

Before you get started

Before you decide to use any of the above tools, consider the weaknesses of artificial intelligence and the ways in which the bot may negatively impact your values.

Privacy
Bias
Currency
Specificity
Transparency
Relevance
Sources
Citations
Academic Integrity
Sustainability / Environmental Costs
Authenticity
Accountability

Always consult with your advisors and instructors before using AI in your academic work. Many publishers also have policies around the use of AI in the work they accept for publication.

his flowchart was designed with ChatGPT in mind, but provides a helpful model for working with any generative AI.

Using ChatGPT with Caution

For detailed discussion on the weaknesses and pitfalls of AI, see https://writingcenter.unc.edu/tips-and-tools/generative-ai-in-academic-writing/

For more information on environmental costs of AI, see:

Chen, Sophia. "How much energy will AI really consume? The good, the bad and the unknown." Nature 639, no. 8053 (2025): 22-24.

"Using ChatGPT with Caution" and "Prompt Examples and their Functions" are adapted from the AI Literacy Module developed by the Rush University's Center for Teaching Excellence and Innovation. They are used under a CC BY-NC-SA License.

Beware of AI Hallucinations

"These systems [artificial intelligence tools] use mathematical probabilities to guess the best response, not a strict set of rules defined by human engineers. So they make a certain number of mistakes. 'Despite our best efforts, they will always hallucinate,' said Amr Awadallah, the chief executive of Vectara, a start-up that builds A.I. tools for businesses, and a former Google executive. 'That will never go away.'"

Cade Metz and Karen Weise, “A.I. Is Getting More Powerful, but Its Hallucinations Are Getting Worse,” The New York Times, May 5, 2025, sec. Technology, https://www.nytimes.com/2025/05/05/technology/ai-hallucinations-chatgpt-google.html.

"Chatbots make things up constantly. They’re also designed to be friendly and agreeable so you’ll spend more time using them. The combination sometimes results in obsequious nonsense, like our Washington Post colleague who found OpenAI’s ChatGPT invented passages from her own published columns and fabricated why that was happening." Hallucinations are a part of how these programs are trained to keep you chatting and using them. They will support your ideas to a fault.

Ovide, Shira, and Heather Kelly. “6 Tips to Avoid Using AI Chatbots All Wrong.” The Washington Post, June 20, 2025, sec. Analysis. https://www.washingtonpost.com/technology/2025/06/20/ai-chatbot-mistakes-users/.

Beware of Lack of Privacy

Ways in which using an artificial intelligence might invade your privacy or those of who you're researching/asking about:

Meta AI (built into Facebook and Instagram as well as Meta's other products) will share your chat transcript publicly with other users.
Anthropic Claude will store your conversations for up to ten years if you use the thumb up or thumb down feature.
If you're using a free version, likely your data is being used in some way. Make sure you check terms of service documentation.
OpenAI is marketing its product to universities, and advocating for a memory feature: "which [would retain] and can refer to previous interactions with a user, would help ChatGPT tailor its responses to students over time and make the A.I. “more valuable as you grow and learn," but would it actually allow a company more access to you in order to sell things to you or sell things about you?

A good rule of thumb: Before confiding in chatbots, imagine how you’d feel if the information you’re typing were subpoenaed or leaked publicly.

Ovide, Shira, and Heather Kelly. “6 Tips to Avoid Using AI Chatbots All Wrong.” The Washington Post, June 20, 2025, sec. Analysis. https://www.washingtonpost.com/technology/2025/06/20/ai-chatbot-mistakes-users/.

Singer, Natasha SingerNatasha, Who Covers Tech in Schools, reported from New York, and San Diego. “Welcome to Campus. Here’s Your ChatGPT.” The New York Times, June 7, 2025, sec. Technology. https://www.nytimes.com/2025/06/07/technology/chatgpt-openai-colleges.html.