Skip to content

A list of useful AI tools for your thesis

There are thousands of tools, but for a thesis you only need a few well-chosen ones. Here they are, grouped by task.

All articlesPublished October 20, 20264 min read

Lists of "the 100 best AI tools" don't help much when you're writing a thesis. You need to know what to reach for when you're looking for literature, when you're drowning in thirty papers, or when you want to check your English abstract. So we've grouped these tools by task. We don't list prices or exact features because they change quickly, so check those on each tool's website.

One important thing first: before you start working with AI, find out what your school's rules say about it and what your supervisor thinks. There's more on that in How to use AI responsibly on your thesis.

Finding literature

Semantic Scholar is an academic search engine that uses AI to sort and summarise papers. It's handy when you want to see quickly which papers are influential in an area and what cites them.

Elicit is a research assistant. You ask a question, it finds papers and helps you pull out details from them in a clear form, such as methods or results. It's especially useful for getting your first bearings in a topic.

Consensus answers questions based on scientific papers. It works well for questions like "Does X affect Y?", when you want to see what research says. Always check the answers against the original papers.

Keep in mind that these tools mostly work with English-language literature. For Czech and Slovak topics, also search your library catalogue and theses.cz.

Mapping connections between papers

When you have one good paper and want to find similar ones, try Connected Papers. It shows a graph of related work, so you can quickly see what belongs to the topic and which older texts are key. ResearchRabbit works in a similar way and helps you add related papers to your collections.

Working with what you've already read

NotebookLM from Google works with sources you upload yourself. You can ask about the contents of your PDFs and notes, and the answers are based on those materials, so you can trace them back to the text. It's practical when you want to recall exactly what an author claimed. It doesn't replace reading, though: a summary won't tell you whether a paper's method is sound.

General chatbots such as ChatGPT, Claude or Gemini are good for explaining a difficult concept, brainstorming, or commenting on a text you wrote. They're not suited to finding citations, because they can make up sources that don't exist.

Managing sources and citations

Zotero isn't an AI tool, but it belongs on this list because it saves a lot of work. It saves papers from your browser, keeps them together with your notes and inserts citations into Word in various citation styles. A Czech alternative is Citace PRO, which some schools provide. Either way, check the generated citations against your school's guide.

Translating and proofreading

DeepL is a translator that's useful for a first version of an English abstract or for understanding a paper in another language. Check technical terms in a dictionary or in papers from your field, because a translator may not choose the right one.

Grammarly helps with the grammar and style of English texts. For Czech, the spell check in Word will do, and when in doubt there's the Internet Language Reference Book, which isn't AI but is reliable.

Transcribing interviews

If you're doing qualitative research, transcribing interviews takes a lot of time. The Whisper model converts speech to text and can run on your own computer, though setting it up takes some technical skill. Always check the transcript by listening, especially names and technical terms. Before you send a recording to any online service, consider whether your respondents agreed to it and whether you're allowed to do that with such data.

A few rules that apply to everything

  • Open and read every source. An AI summary can be inaccurate, and a paper may say something other than it seems.
  • Don't casually upload sensitive data, such as non-anonymised interviews or a company's internal materials, to online tools.
  • Keep a note of what you used and what for. When your supervisor asks, or when you have to state it in your thesis, you'll have it to hand.

Where to start if you don't have a topic yet

All the tools above assume you know what you're looking for. If you're still choosing a topic, you can start with Thesios.ai. Based on your school, field and interests, it suggests 10 narrow topics for free and, for a fee (or through your school, if it pays for you), a detailed proposal with an outline and literature from Google Scholar and Crossref. It's a starting point: it won't write your thesis, and you have to check what it produces.

Common questions

Am I allowed to use these tools for my thesis?

That depends on your school's rules and what you agree with your supervisor. Finding literature or checking typos tends to be less of an issue than generating text, but always check first and state your AI use in the thesis if required.

Which tool should I try first?

Start with source management, such as Zotero, because it saves you work the whole way through. Add one search tool, for example Semantic Scholar, and try others only when you have a specific need.

Can I trust citations a chatbot gives me?

Not without checking. Chatbots sometimes invent papers that look believable but do not exist. Find every source in a database or by its DOI, and read it.

10 free topics for your field

Enter your field and interests and you'll see the first suggestions within moments. No payment, no sign-up.

Find 10 free topics