Overview of AI-driven research tools

  • AI is a rapidly evolving field, with tools and technologies advancing quickly. This page was last updated in February 2025. We make regular updates to ensure the information remains current. For the most up-to-date insights, you can also visit the Generative AI Product Tracker by Ithaka S+R.
  • If you know of other useful tools or have suggestions for improving this page, please contact the University Library at dirk.schoenaers@uhasselt.be.
  • Disclaimer | The lists below are not an endorsement by Hasselt University or its staff. They are provided solely to illustrate the potential and limitations of current AI tools for academic research.

AI@UHasselt

  • AI@UHasselt | for researchers and companies | AI research, communities, news, and events at UHasselt.
  • RDM Gen AI page | for research and teaching staff | general framework on the use of (gen) AI at UHasselt and guidelines towards responsible use in research.
  • Effective and responsible use of AI in your studies page (Dutch version: here) | for students | practical tips and guidance on leveraging AI tools to support your studies, while avoiding common pitfalls.

Key considerations

  • AI platforms act as support tools. Researchers are ultimately responsible for their use of AI tools and their output. 
  • The training data behind these tools is often unclear, complicating source verification and potentially impacting licensing | copyright (cf. below).
  • Always verify results thoroughly, recognize limitations and biases, and privacy issues (cf. below).
  • Be transparent about how generative AI is used in your research. Some basic guidelines for citing AI in research in APA style.

Legal & ethical responsibilities

When using online tools (gen AI, but also online translation platforms or plagiarism checkers) researchers should ensure compliance with data protection and intellectual property laws, ethical standards, and the attribution of AI-assisted work. Failure to do so may lead to unintended legal, ethical, or academic repercussions. If unsure, do not use the tool or contact ai@uhasselt.be.

  • Thoroughly review the licensing agreements, especially for free versions of (AI) tools and products that do not fall under a university-wide license. Pay close attention to terms governing usage rights, data privacy, and intellectual property.
  • Understand if and how your input data be stored and (re)used for other purposes, potentially raising concerns about confidentiality, data misuse, and intellectual property rights. 
    More on information and data security. 
  • Never share personal, sensitive, or confidential information with AI tools, including: 
    • Any information protected under data protection laws (e.g. GDPR): names, addresses, or identification numbers.
    • Data with ethical concerns about misuse or group harm.
    • Data essential for future valorization or protected by intellectual property law.
    • Proprietary or commercially sensitive information.
    • Data subject to contracts or funding restrictions on sharing.
    • Information and data shared under an NDA, unless stated otherwise in the NDA.
    • Copyrighted or database-protected data, unless permitted by the owner
  • Keep in mind that research results generated entirely by AI are unlikely to qualify for protection under intellectual property rights. However, if a research result is produced using AI but includes significant input from the researcher (e.g. providing highly specific instructions to a gen AI tool), it may be eligible for such protection. Current legislation and case law do not yet offer a definitive answer on this matter.
Gen AI in research

Generalist models & tools

General-purpose Large Language Models like ChatGPT are trained on broad datasets. Their outputs can be impressive in breadth, but commonly lack the precision and domain-specific expertise that many research tasks require.

  • LLMs are versatile, but struggle with highly detailed subject-specific knowledge, complex methodologies, and source verification.
  • The training data may be over a year old.  Models and tools without real-time internet access will deliver outdated results.
  • LLMs are prone to bias based on their training data and may generate fabricated or incorrect information as fact ('hallucinations').
  • Training materials do not (always) include academic sources that are not publicly available.

As a result, it remains crucial to verify information with reliable academic sources, just as you would with any general online search.

  • Training LLMs: Paid versions of tools like ChatGPT provide options for customization and training. When you train your own model, pay careful attention to the training data, especially personal data and its (pseudo-)anonymization.
  • Open-source LLMs, like Llama and those available through  Hugging Face, can be run locally on laptops or cloud instances, which often reduces privacy concerns. However, these models present unique challenges, such as the technical expertise needed for installation and operation.
List of generalist LLMs | Tools

AI-driven research tools

AI-driven research tools utilize artificial intelligence (AI) to support and enhance various aspects of the research process. These tools apply techniques such as machine learning, natural language processing, and data analysis to automate or optimize tasks such as data collection, processing, analysis, and interpretation.

AI-driven research tools

  • are designed to filter, organize, and analyze scholarly articles more effectively than general LLMs, offering more reliable literature reviews, source-tracking, and | or advanced data analysis capabilities tailored to academic workflows.
  • are often further trained with subject-specific or research-oriented datasets.
  • more commonly employ mechanisms designed to minimize hallucinations and reduce bias.

Nonetheless, it remains essential to check results and be transparent about your use of AI tools. 

These tools can offer assistance in several fields:

List of AI research tools

Basic guidelines for prompting

Prompts are open-ended instructions intended to guide responses, typically used in AI models and expressed in natural language. They promote the generation of extended answers, inviting deeper exploration. This sets them apart from traditional queries or specific and targeted search commands. Queries often involve a combination of keywords, Boolean operators (AND, OR, NOT), truncation symbols (*, ?), and string search ("...") and focus on directly retrieving relevant data or information from a system or database.

  • Be clear and specific about your topic and research discipline, the specific task, and the expected output.
    Example: 'I am working on a dissertation in environmental economics on renewable energy policies in the EU. Can you suggest recent papers on this topic?'
  • Include keywords and phrases to narrow down the search.
    Example: 'Look for research articles on 'gene therapy for cystic fibrosis using CRISPR-Cas9'.'
  • Include carefully selected examples to get more relevant results.
    Example: '[...] For example, I am looking for results similar to: ‘A 2021 study in UK high schools showed that implementing mindfulness programs reduced anxiety scores by 30%.’
  • Break down complex queries into smaller, manageable parts.
    Example: 'Could you help me formulate a hypothesis for a study on the effects of online learning on student engagement?.'
  • Explicitly request academic sources or citation support
    Example: 'Please list three key systematic reviews on telemedicine’s effect on patient outcomes. Include citation links where possible, ideally from high-impact journals.'
  • Ask for multiple perspectives or comparisons
    Example: 'Provide a comparison of recent studies on 'supply chain resilience' in the manufacturing vs. retail sectors. Focus on how each sector addresses disruptions.'
  • Iterate for better results  |  Refine or rephrase your query if the initial answer does not meet your expectations.
    Example: 'Your initial summary on sustainable agriculture practices was useful, but could you include a comparison of practices used in temperate vs. tropical climates?'

! For reliable citation data and finding academic publications, traditional scholarly resources like Web of Science or the UHasselt Discovery service remain the most dependable options.

You can find a more detailed overview of specific prompting techniques (e.g. prompt chaining) in this guide.

Tools for detecting AI use

While AI detection tools may seem useful, their judgment is generally not very reliable. Some tools, like OpenAI's Text Classifier, focus specifically on detecting text generated by their own models (e.g., GPT).

Tools like Turnitin AI Checker, GPTZero, and Copyleaks offer a higher level of reliability due to their extensive databases and more advanced algorithms. However, false positives and negatives remain a common issue, especially if the text is highly structured | written in a style that resembles typical AI outputs (formulaic) or when AI-generated text has been heavily edited by a human.

As AI technology evolves, so too will the reliability of detection tools, but for now, their use should always be supplemented with careful human review to ensure accuracy and fairness.

Plagiarism checkers

Using Gen AI as a writing assistant significantly increases the risk of plagiarism. Online checkers have become popular tools for detecting plagiarized content. 

The reliability of plagiarism checkers depends on several factors, such as

  • size and contents of the database
  • algorithm accuracy, and
  • handling of citations.

Larger databases and advanced algorithms can detect paraphrasing and subtle plagiarism. Nonetheless,  false positives may occur, e.g. when  common phrases or properly cited quotes are flagged as plagiarism. 

! User privacy and security: be aware that some online checkers retain user submissions, which could lead to future plagiarism flags or privacy concerns.

List of plagiarism checkers

Basic guidelines for citation

Adding AI tools to the reference list of a publication provides a direct way to cite the technologies used. Below you will find some basic guidelines for citing AI in APA style (7th Edition). 

In-text citation of an AI tool:

Author/company or creator of the tool (year of the version used)

  • OpenAI (2024) states that the use of AI tools in education offers many advantages.
  • The use of AI tools in education offers many advantages (OpenAI, 2024).

 

Reference list entry for an AI tool:

The full reference to the AI tool used should be included in the reference list as follows:

Author/company or creator of the tool. (Year of the version used). Name of the tool or model (release date/version number/version name) [Type of AI model]. Retrieved on  month day, year, from URL of the tool's website

  • OpenAI. (2024). ChatGPT (April 29, 2024) [Large language model]. Retrieved on June 25, 2024, from https://chatgpt.com/
  • Google DeepMind. (2023). Gemini (December 6, 2023) [Large language model]. Retrieved on December 14, 2023, from https://gemini.google.com/
  • OpenAI. (2023). DALL-E (Version 3) [Text-to-image model]. Retrieved from https://bing.com/

 

The citation style (APA, Chicago, Vancouver...)  may vary depending on the requirements of each assignment, academic journal, or funding body.

Generalist models & tools

Generalist AI models can generate text, summarize information, answer questions, and support writing tasks. Tools like Gemini and Edge Copilot are more effective for real-time searches and web-based tasks, while open-source models like Llama allow for customization to specific needs. DeepSeek's R1 model matches or surpasses ChatGPT, MS Copilot, and Gemini in tasks that simulate human language and knowledge. However, its operation under China's regulatory framework may influence content availability and responses on sensitive topics.

In search-based tasks, these models tend to reference broadly accessible general sources rather than scholarly publications, unless specifically prompted to do so. This impacts the accuracy and relevance of their output for academic research, especially on topics needing specialized knowledge or recent data. For reliable academic information, it is recommended to use specialized databases or verified academic publications directly. Any AI outputs should be reviewed carefully.

Model | Tool

Main characteristics

Use & limitations

Free | Paid subscription

Chat GPT
  • capable of drafting text, summarizing information, answering questions, and creating content in various formats.
  • trained on diverse internet sources,.
  • knowledge cutoff  January 2022 (GPT 3.5) | April 2023 (GPT 4) | October 2023 (GPT 4o).
  • ChatGPT Edu (subscription only): 'a version of ChatGPT built for universities to responsibly deploy AI to students, faculty, researchers, and campus operations.'

     
  • useful for brainstorming, refining research questions, search strings, and selection criteria. 
  • can assist in drafting outlines and suggesting improvements for clarity, coherence, and flow.
  • efficient for language assistance, translation, and summarizing, but always cross-check the details.
  • allows uploading files for analysis | processing.

 

  • not suited for bibliographical research.
  • does not automatically include citations or verifiable sources.
  • may produce outdated information and hallucinations.
  • may fabricate bibliographical references. Explicitly ask for direct links to (the full text) of verifiable academic sources. 
  • should not be used to replace domain-specific research expertise. 
  • output must always be critically reviewed against reliable academic sources, modified, and referenced.

Free version

  • access to a version of ChatGPT with an earlier knowledge cutoff.
  • limited access to advanced features.
  • may have restrictions on usage frequency or prompt length.

 

Paid subscription 

  • access to latest models and premium features.
  • faster response times and more stable service.

Claude
  • designed for drafting text, summarizing, answering questions, and creating content.
  • trained on diverse datasets. 
  • Claude 3 knowledge cutoff date: August 2023.
  • useful for brainstorming, developing content ideas, refining complex concepts, coding.
  • effective in summarizing and identifying key elements in academic frameworks (e.g. PICO).
  • can provide language support, including translation and grammar improvement.
  • allows file upload.
  • user input not used for training purpose, unless user opts in.
  • improved safety.

 

  • does not automatically include citations or sourcing. Cannot directly link to (the full text of) verifiable academic sources.
  • responses can sometimes be over-generalized or contain outdated information.
  • information needs to be cross-checked against verified resources.
  • not designed to replace subject-specific expertise.
  • content must always be reviewed, edited, and properly referenced.

Free Version:

  • access limited to an earlier model with restrictions on capabilities or frequency of use.

 

Paid Subscription:

  • access latest model with improved speed, stability, and access to advanced features.

Gemini
  • integrates Google’s search capabilities with AI-powered conversation to generate contextually accurate and real-time answers.
  • designed for both text generation and factual searches.
  • Gemini Pro knowledge cutoff date: April 2023.

 
  • useful for search-based queries. Gemini draws from Google's live search engine, ensuring current data.
  • effective in summarizing content.
  • returns source references.
  • offers fact-checking (Double-check response).
  • useful for language support, translation, and improving written content.
  • more robust image search.

 

  • main strength is search-based responses. Creative or subjective tasks (e.g., content brainstorming, in-depth analysis) may be limited.
  • information from real-time sources may lack deep context, especially in areas requiring domain-specific expertise.

Free Version:

  • may have limitations in response depth or frequency of queries.
  • search functionality remains strong.

 

Paid Subscription:

  • priority access to advanced features, faster response times, and  improved accuracy.

Llama
  • series of LLMss developed by Meta (Facebook) for e.g. text generation, summarization, translation, and question answering.
  • primarily designed as an open-source AI model, allowing fine-tuning for specific tasks or applications.
  • Llama 3 7B has a knowledge cutoff from March 2023; for Llama 3 80B this is December 2023.
  • useful for research and experimentation in AI and NLP due to its open-source nature, enabling developers to customize it for specific use cases.
  • Useful for text-based tasks, such as content generation, summarization, and translation.
  • effective for projects where fine-tuning is required to adapt to specialized tasks.
  • can assist in language processing, such as parsing large datasets, summarizing texts...

 

  • often requires fine-tuning to reach its full potential.
  • less suited for general users who need a plug-and-play AI solution.
  • training and deployment of larger models can require significant computational resources.
  • still prone to producing incorrect or outdated information. The generated content needs verification.

Free Version:

  • anyone can download and work with the LlamA models for free.

 

Paid Subscription:

  • no paid subscription model 
  • commercial applications based on LlamA (if developed by third parties) may offer enhanced features or subscription-based services.

Edge Copilot
  • embedded within Microsoft’s Edge browser.
  • can offer contextual information, answer questions, summarize content, and interact with web pages.
  • can assist with basic writing tasks.
  • generates images.
  • Copilot for Microsoft 365 (subscription only).

 

 
  • best suited for general web tasks.
  • useful for gathering information quickly and getting clarifications while browsing.
  • basic content generation and writing assistance.

 

  • primarily tied to web-based tasks.
  • accuracy may vary. Checking and cross-referencing is essential.
  • may produce plausible but incorrect information.
  • domain-specific research requires more specialized tools.

Free Version:

  • available within the Microsoft Edge browser for all users.

 

Paid Subscription:

  • Copilot Pro offers priority access to latest models.
  • higher usage limits.

DeepSeek
  • developed by Deepseek AI, a China-based AI company.
  • offers multiple models, including Deepseek Chat (for general AI interactions) and Deepseek Coder (optimized for code generation).
  • trained on multilingual datasets, with a strong emphasis on Chinese and English.
  • claims higher efficiency in computational resource use compared to models like GPT-4 and Gemini.
  • good for factual accuracy due to retrieval-augmented generation. Nonetheless, output should be verified against reliable resources (see below).
  • strong performance in code generation (Deepseek Coder).
  • open-source availability allows for custom fine-tuning.
  • can handle complex problem-solving, making it suitable for mathematical and scientific tasks.

  • less optimized for creative writing tasks.
  • some concerns about content moderation and censorship due to Chinese regulatory influence.

 

Free version:

  • available for basic usage without cost.

AI research tools

AI tools specifically developed for research, are built with a sharper focus on academic tasks. They are designed to filter, organize, and analyze scholarly articles more effectively, offering more reliable literature reviews, source-tracking, and advanced data analysis capabilities tailored to academic workflows. These tools are often further trained with subject-specific or research-oriented datasets. In addition, they commonly employ mechanisms designed to minimize hallucinations and reduce bias.

The lists below are not meant to be exhaustive, but rather aim to give an overview of the possibilities and limitations of AI tools tailored for or used in academic research. Various tools can provide support in the following areas:

Literature search & Discovery

The main purpose of the following tools is to help you search for research articles and identify trends. Caution is advised, since occasional citation errors may occur. These tools often use the datasets of repositories such as Semantic Scholar, OpenAlex, Crossref, and | or Pubmed to source (bibliographical) information. A general rule of thumb is to be wary of any platform that does not clearly indicate where its articles come from. For now, these tools work best as a supplement to traditional database searches. 

Several academic publishers and commercial companies are developing their own AI research assistants, integrated into their databases. Examples include Clarivate | ProQuest and Elsevier (Scopus). These tools are often only available through a paid license. They are typically designed to query the respective databases using natural language, summarize the most relevant or highly cited results (or more accurately their metadata), and create visualizations of relationships between publications. UHasselt currently participates in the beta programme for the JSTOR interactive research tool. Legal databases, such as Strada lex (GenIA-L) and LexNow (LexGAIN), have also developed their own AI-powered assistants or are doing so (monKEY). We will keep you updated on further developments.

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Semantic scholar
  • AI-driven discovery tool with access to over 200 million academic papers in STEM and HSS fields.
  • also includes preprints.
  • features include summaries and auto-highlights. 
  • useful for discovering new papers in STM fields.
  • suitable for quick exploration of key points in literature. 
  • offers citation support.

 

  • Summary TLDR feature limited to certain disciplines. 

Free version only

Perplexity
  • search-based AI assistant.
  • combines LLM capabilities with real-time web searches.
  • provides linked citations for traceable information.

 
  • useful for news, current events, but also emerging research topics because of live searches.
  • useful for quick summaries and following up on primary sources.
  • can assist with general content creation, answering questions, summarizing, and basic brainstorming tasks.
  • linked citations improve the verifiability and traceability of the information.

 

  • the quality and relevance of results can vary. Cross checking for accuracy remains advisable.
  • tends to focus more on retrieving search-driven facts rather than nuanced reasoning or abstract content creation.
  • search results may not capture the necessary depth for highly specialized topics.

Free Version:

  • Unlimited Quick searches
  • 5 Pro Searches per day
  • Standard Perplexity AI Model optimized for speed and quality

 

Paid Subscription:

  • Unlimited Quick searches.
  • 300+ Pro searches per day.
  • Select your preferred AI Model. 
  • Upload and analyze unlimited files.
  • Visualize answers.

Elicit
  • tool for automating research workflows
  • features for finding and summarizing academic papers, data extraction, and thematic synthesis.
  • claims to reduce inaccuracies by: fine-tuning models for specific tasks, displaying only verified papers, and making it easy to access the original source for cross-checking answers.
  • source: Semantic Scholar.

 

 
  • suitable for conducting structured literature reviews.
  • useful for summarizing research and data extraction.
  • automatically organizes information from multiple papers into comparison tables and identifies common themes.
  • users can upload their own documents for analysis.
  • best for empirical research.

 

  • may oversimplify abstract or theoretical work and data extraction sometimes lacks prioritization of the most relevant content​.
  • requires cross-checking results.

Free version

  • free credits, with pay-as-you-go options after the limit.

 

Paid subscription

  • enterprise-level subscriptions available for teams.

Scinapse
  • AI-powered academic search engine offering over 48,000 journals
  • similar to Elicit: helps find relevant papers, apply filters, and analyze research insights.
  • provides trend analysis and researcher discovery features.
  • sources: PubMed, OpenAlex, Semantic Scholar.
  • useful for tailored research recommendations, author tracking, and research trends.
  • advanced filters and Scinapse's 'Expert Finder' make identifying leading researchers and emerging trends easier. 

 

  • limited full-text availability and integration options.
  • pro version required for in-depth trend analysis and expert search capabilities.

Free version

  • access to standard paper search, journal browsing, and basic recommendations.

 

Paid subscription

  • offers features like Research Intelligence and Expert Finder​.

Litmaps
  • literature mapping and discovery tool that
  • visualizes connections between research papers based on citation networks.
  • features include Seed Maps, real-time alerts, and customizable monitoring for new publications.
  • sources: Crossref, other OA metadata repositories.
  • useful for tracking ongoing research, identifying gaps, and visualizing paper relationships.

 

  • limited in-depth content analysis.
  • more useful for discovering and monitoring than for detailed review. 

Free version

  • two litmaps with basic version.

 

Paid subscription

  • unlimited Litmaps.
  • advanced filters.
  • monitoring alerts.

Keenious

 
  • analyzes your document to recommend relevant academic research.
  • searches based on the content of your text, rather than just keywords, offering relevant articles across various disciplines.
  • source: OpenAlex.
  • useful to enhance literature reviews, discover cross-disciplinary research, or explore new topics.
  • cross-language search finds research in different languages.
  • generates citations in multiple styles and integrates with reference management systems.
  • provides smart citation management and allows integration with tools like Google Docs and Microsoft Word.

 

  • free plan limits the length of documents analyzed and daily searches.

Free version

  • limited to analyzing short documents.
  • fewer recommendations.
  • one cross-language search per day.

 

Paid subscription

  • expanded results.
  • unlimited analysis.
  • smart bookmarks.
  • advanced filters.

ResearchRabbit
  • discovery tool offering paper and author searches with network visualization.
  • enables real-time updates for collections and collaboration.
  • sources: PubMed, Semantic Scholar.

 
  • useful for visualizing research collections and tracking authorship trends.
  • integrates with Zotero for improved citation management.

 

  • limited for novel research; focuses on existing citation networks. 
  • focuses on journal articles and may miss non-journal resources. 

Free version only

Inciteful
  • citation-based tool for visualizing research connections and academic influence through citation impact analysis. 
  • sources: OpenAlex, Semantic Scholar, Crossref, OpenCitations.
  • suitable for visualizing citation trends and academic networks.
  • helpful in revealing connections and overlooked studies.

 

  • limited primarily to journal articles.
  • relies on citation data, which may overlook recent or interdisciplinary connections. 

 

 

Free version only

Connected papers
  • visual graph exploration of academic papers, showing relationships based on citations.
  • source: Semantic Scholar.
  • useful for discovering papers connected to a core paper.

 

  • depends on Semantic Scholar database, which may not include all papers, especially those outside of scientific fields.
  • free plan limits users to five graphs monthly​

Free version

  • all features, up to 5 graphs per month.

 

Paid subscription

  • academic and business plans offering unlimited graphs.

Open Knowledge Maps
  • generates a map of a research topic to aid in locating documents and identifying key concept.
  • aggregates related literature into clusters based on topic similarity.
  • provides a quick visual overview.
  • source: PubMed, BASE.
  • useful for quick overview of a research field, identifying key papers, and exploring interdisciplinary connections.

 

  • restricted to 100 documents per map.
  • quality of clustering may vary based on metadata accuracy.

Free version only

Jane (Journal/Author Name Estimator)
  • helps researchers find relevant journals, authors, or articles by analyzing a manuscript's title or abstract.
  • sources: PubMed, MEDLINE, DOAJ.
  • useful for identifying journals for submission or finding relevant authors.

 

  • results may include lower-quality journals, as some predatory journals are indexed in PubMed. Users need to evaluate journal suitability independently​.
  • best suited for STEM fields, with limited scope for humanities. 

Free version only

Iris.ai
  • simplifies research discovery and mapping by using semantic analysis rather than keyword search.
  • automates literature reviews and categorizes research articles.
  • source: OA articles via CORE and PubMed.
  • useful for finding connections between different fields, and expanding literature searches.
  • suitable for large-scale literature reviews.

 

  • limited for deep analysis or specialized fields.

Free version

  • basic search and discovery features.

 

Paid subscription

  • advanced features include automated literature reviews, categorization, and full access to workspace tools​.

Systematic review & Literature screening

Using software and tools with machine learning algorithms and text-mining techniques to handle large datasets can automate key aspects of systematic reviews, such as screening, data extraction, and synthesis. However, automation comes with risks, like the potential to miss nuanced details that require expert human judgment. Maintaining transparency is essential; protocols or reports should clearly document the tools used, the criteria for their application, and any decisions shaped by automated methods.

For more information on tools that can be used for systematic reviews, please consult our Systematic and Systematized Reviews page (under 'Tools & Software').

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Rayyan
  • offers AI-powered tools for study screening, deduplication, and collaboration.
  • integrates with reference managers like Mendeley.
  • designed to help researchers collaborate on screening literature for systematic reviews.
  • enables rapid title and abstract screening, coding, and labeling of studies for inclusion or exclusion.

 

  • limited to the screening process.
  • does not assist with later stages like data extraction or synthesis.
  • free version is limited to three active reviews and standard support.
  • advanced features (e.g., relevance ranking, PICO highlighting) require a subscription.

Free version

  • basic version with three active reviews.

 

Paid subscription

  • offers unlimited reviews, priority support, and advanced tools​.

DistillerSR
  • automates and manages all stages of systematic reviews.
  • offers AI-driven tools for screening, data extraction, and report generation. 
  • useful for conducting systematic reviews, providing tools for collaboration and organization.

 

  • primarily suited to large-scale or enterprise-level projects.
  • subscription costs can be high.

 

Free version

  • free trial.

 

Paid subscription

  • pricing based on institutional or enterprise needs.

Writing assistance

AI writing assistants can enhance efficiency, clarity, and language precision in academic work. However, they often lack the depth needed for complex topics, may introduce factual errors, and can occasionally risk unintentional plagiarism. Careful review is crucial to maintain accuracy and integrity.

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Grammarly
  • comprehensive AI-powered writing assistant.
  • designed to improve grammar, punctuation, clarity, engagement, and delivery in various types of written content.
  • widely used for both academic and professional purposes.
  • core features include grammar, punctuation, and style correction, plagiarism detection, AI-powered suggestions for improved readability.
  • useful for refining academic writing, ensuring clear and professional language, and avoiding grammatical errors. 
  • reduces the time spent on editing and proofreading.
  • Grammarly Premium supports citation integrity with plagiarism detection.

 

  • plagiarism detection may not capture all scholarly sources and cannot replace human review for nuanced academic content.

Free version

  •  basic features.

 

Paid subscription

  • offers advanced features, including plagiarism detection. UHasselt students and staff can request a premium account at no personal charge and login with their UHasselt credentials. 

DeepL write
  • offers grammar corrections, rephrasing suggestions, and style adjustments.
  • multiple languages, including English, German, French, and Spanish.

 
  • useful for (non-native English speakers) writing academic content.

 

  • current beta lacks advanced customization like glossary features.

Free version

  • free beta version with limited text improvements.

 

Paid subscription

  • offers unlimited improvements.

Jenni.ai
  • offers AI-powered support for academic writing
  • topic suggestions.
  • in-text citations.
  • paraphrasing.
  • PDF reading.
  • provides real-time suggestions as users write.
  • useful for speeding up writing with auto-complete and suggested citations.
  • useful for idea generation, structuring, and citation management.
  • offers a 'line-by-line' writing aid.
  • useful for focused academic work but may feel slower for extensive content creation.

 

  • limited content tone and format adjustments.
  • citation accuracy and paraphrasing reliability may vary.
  • requires manual verification. 

Free version

  • basic access.
  • limited to 200 AI words per day.
  • AI autocomplete.
  • journal & web citations.
  • unlimited PDF uploads.

 

Paid subscription

  • gives access to unlimited AI words per day.
  • access to latest features.

Quillbot
  • provides  paraphrasing, summarizing, grammar and plagiarism checking, citation generation, and translation.
  • integrates with Google Docs, Microsoft Word, and Chrome extensions.
  • useful for improving writing quality.
  • useful for managing citations.

 

  • not ideal for detailed academic analysis.
  • free plan limits paraphrasing to 125 words.
  • plagiarism checker scans only 20 pages per month

Free version

  • offers basic paraphrasing (up to 125 words), summarizing (up to 1200 words), and limited grammar check.

 

Paid subscription

  • offers more features such as advanced paraphrasing, grammar checks, summary support.
  • includes plagiarism check against online sources.
  • offers access to co-writing platform.

Writefull
  • tailored for academic writing.
  • provides grammar checks, paraphrasing, title and abstract generation.
  • includes GPT detector to identify AI-generated content.
  • enhances clarity, accuracy, and adherence to scholarly standards in writing.
  • integrated with platforms like Microsoft Word and Overleaf,
  • offer in-text feedback.
  • focus on academic language.

 

  • AI suggestions occasionally misaligned with nuanced meaning.
  • false positives | negatives in GPT detection.
  • trained on scientific texts, less suitable for general content. 

Free version

  • basic features with limited access to language suggestions.

 

Paid subscription

  • full access to all features and comprehensive language suggestions​.

SciSpace | Typeset.io
  • writing and formatting tool for academic papers.
  • helps format and structure writing. 
  • suitable for formatting manuscripts according to journal guidelines.

 

  • not suited for content generation or in-depth analysis. 

Free version

  • basic access.

 

Paid subscription

  • offers more formatting and editing tools.

NotionAI
  • a flexible, all-in-one workspace for note-taking, project management, and collaboration.
  • supports a variety of templates and integrations (e.g., with Slack, Zapier) and customizable workflows.
  • useful for tasks like note-taking, project management, and document collaboration.

 

  • the free plan has a 5 MB file upload limit and allows only up to 5 guests for collaboration. 

Free version

  • Unlimited pages and blocks for personal use, with limited upload and collaboration (up to 5 guests).

 

Paid subscription

  • offer larger file uploads, priority support, and advanced collaboration.

Paperpal Copilot
  • AI writing assistant designed for academics
  • provides real-time grammar and language suggestions, academic translations, plagiarism checks, and technical compliance checks for manuscript submission.
  • useful to improve manuscript quality, clarity, and compliance with journal standards.
  • offers plagiarism checks.
  • translation in over 30 languages, optimized for scientific writing.
  • offers subject-specific suggestions for over 1,300 academic fields.

 

  • free plan is limited to 200 language corrections and 5 AI features per day.

Free version

  • limited language suggestions, AI uses and plagiarism check. 

 

Paid subscription

  • unlimited corrections, AI tools, plagiarism checks, and manuscript readiness checks​.

Grantable 
  • assists users in drafting proposals by analyzing previous submissions and generating relevant content for new applications. 
  • provides AI-driven drafting and revision tools
  • speeds up the grant writing process.
  • paid version supports multiple users, enabling teams to work together on grant proposals

 

  • effectiveness depends on the quality of the input data.
  • requires manual review to ensure compliance with specific requirements. 
  • free tier limits users to 10 AI actions per month

Free version

  • Includes one user seat.
  • unlimited document creation.
  • up to 10 AI actions per month.

 

Paid subscription

  • various plans. 100 or unlimited AI interactions per month.

Granted AI 
  • AI assistant for identifying funding opportunities and writing grants. 
  • offers a wide range of specialized writing modells, targeting specific grants and funding agencies.
  • users can select the best-generated content, rearrange it, add or delete as needed.
  • provides constructive feedback.

 

  • limited in detailed content creation. 

Free version

  • free trial.

 

Paid subscription

  • various subscription plans tailored to different needs and organizational sizes.

Summarization & Knowledge extraction

Tools for summarizing research papers or extracting key information from literature by quickly identifying main ideas, concepts, and findings. However, caution should be exercised when using such tools, as they may not always accurately capture nuances, context, or critical details, potentially leading to oversimplified interpretations or misrepresentations of the original research. Always verify the results with a careful review of the full text.

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Scholarcy
  • transforms complex documents into concise, interactive summarie.
  • highlights key information.
  • useful for quick summarization and quickly grasping the core content of articles.
  • creates interactive flashcards from articles.
  • extracts cited references and attempts to link to open-access versions when available.
  • integrates with reference management software.

 

  • may not always capture the nuances of highly complex material. 
  • some features limited in the free version.

Free version

  • basic access.
  • generate up to three summaries per day.

 

Paid subscription

  • offers unlimited usage.
  • enhanced features like saving flashcards, taking notes, highlighting text.

Humata AI
  • enables users to upload documents and interact with them through natural language queries.
  • facilitates tasks such as summarization, question answering, and information extraction.
  • suitable for document-specific queries.
  • useful in summarizing and generating content from existing files. 
  • can synthesize information from multiple documents simultaneously.

 

  • accuracy of outputs may vary depending on the complexity and quality of the input documents.
  • users need to verify the information extracted, especially when dealing with highly specialized or technical content.

Free version

  • basic features with a limit of 60 pages and up to 10 answers per month.

 

Paid subscription

  • various plans supporting larger amounts of pages and more document interaction features.

Consensus building, Argumentation & Insights

Tools that help gather insights and build consensus on research questions or findings by aggregating diverse perspectives and analyzing patterns in academic literature. However, caution should be taken as these tools rely on algorithms and databases that may not fully capture the complexity or evolving nature of research topics, potentially leading to biased or incomplete conclusions. It remains essential to cross-check the findings with original sources and expert evaluation.

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Consensus
  • AI-powered scientific search engine that summarizes areas of consensus from academic research.
  • source: Semantic Scholar.
  • offers a visual representation of the scientific consensus on a given topic, aiding users in quickly gauging the prevailing viewpoints within the research community.
  • Extracts key information from individual studies enabling users to assess the quality and relevance of research papers efficiently. 

 

Free version

  • basic access.

 

Paid subscription

  • offers premium features.

DebateDevil
  • engages users in simulated debates.
  • challenges users' hypotheses by presenting counterarguments, encouraging them to consider alternative perspectives.
  • facilitates exploratory discussions through questioning.
  • provides constructive feedback and positive reinforcement.
  • useful for refining arguments and broadening perspectives.
  • generates evidence-based arguments.

Free version

  • basic access.
  • 10 questions per month.

 

Paid subscription

  • offers more features.

Scite.ai
  • provides context to citations, showing whether they support, mention, or contrast a given paper.
  • users can search for specific claims or statements within the literature, facilitating targeted information retrieval.
  • sources: preprint servers, open access, and agreements with various publishers.
  • aids in conducting comprehensive literature reviews, identifying research gaps, and understanding the context of citations.
  • suitable for verifying citations and understanding their impact.
  • useful in literature reviews to gauge citation strength.

 

  • results need to be critically reviewed for accuracy.

Free version

  • basic access.

 

Paid subscription

  • offers advanced functionalities such as unlimited searches, detailed reports, and integration capabilities.

Interview transcription

AI transcription can save time and effort in tasks like interview analysis, lecture documentation, and data processing. However, accuracy may be inconsistent, especially with specialized terminology, strong accents, or low-quality recordings, often requiring manual corrections. Ethical concerns about data security and privacy are critical, particularly for sensitive research. While efficient, AI transcription should be used with careful consideration of these limitations to ensure reliability and compliance with academic standards.

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Avidnote
  • AI-powered research tool designed to assist researchers in writing, reading, and analyzing academic papers more efficiently.
  • provides templates to streamline the research writing process.
  • helps in drafting and rewriting text, including literature reviews and discussions.
  • offers functionalities to summarize texts, proofread, and identify research gaps.
  • enables users to input data and discover correlations and insights.
  • automatically transcribes interviews, supporting multiple languages while ensuring data privacy.
  • Ideal for researchers, academics, and students seeking to enhance productivity in managing and analyzing research materials.
  • converts recorded interviews into accurate text with minimal manual intervention.
  • allows researchers to edit and refine transcriptions directly within the platform, ensuring accuracy and contextual relevance.

 

  • accuracy may be affected by poor audio quality, overlapping voices, or strong background noise.
  • free plan offers limited transcription hours.

Free version

  • includes 2 hours of AI transcription per month, along with basic AI tools and storage.

 

Paid subscription

  • various plans offering more hours and storage.

Whisper
  • automatic speech recognition (ASR) system developed by OpenAI, trained on a vast dataset to achieve human-level robustness and accuracy in English speech recognition.
  • capable of transcribing multiple languages and translating them into English.
  • designed to handle diverse accents, background noise, and technical language effectively.
  • open-sourced models and inference code are available for developers and researchers.

 
  • suitable for developers building voice interfaces, transcription services, and applications requiring speech-to-text capabilities.

 

  • While highly robust, performance may vary depending on audio quality and language complexity.

Free version

  • developers can utilize the open-source models for free.

 

Paid subscription

  • OpenAI offers an API for Whisper at a fixed cost per minute of audio processed.

Rev
  • Rev offers both AI and human transcription services, catering to various needs for accuracy and speed.
  • human transcription provides accurate transcriptions by professional transcribers.
  • AI transcription delivers faster transcriptions with slightly lower accuracy.
  • integrates with various platforms and tools.
  • suitable for researchers requiring high-accuracy transcriptions.

 

  • human transcription services are more expensive and have longer turnaround times compared to AI transcriptions.

Paid subscription only

  • Rates per minute for AI and human transcription.
  • Subscription plans available for frequent users with additional features.

Trint
  • AI-powered transcription platform that converts audio and video files into text.
  • supports transcription in more than 40 languages.
  • provides an in-browser editor for verifying, editing, and searching transcripts.
  • allows real-time collaboration.
  • integrates with various platforms.
  • useful for researchers needing efficient transcription and editing tools.

 

  • accuracy may vary depending on audio quality; manual review is recommended.

Free version

  • Offers a free trial with limited features.

 

Paid subscription

  • Various plans available.

Sonix
  • automated transcription service that transcribes audio and video files in over 53 languages.
  • additional features like translation and subtitling.
  • includes AI-powered summaries and topic detection.
  • suitable for researchers needing multilingual transcription and translation services.

 

  • automated transcriptions may require manual corrections for complete accuracy.

Free version

  • free trial offers 30 minutes of free transcription.

 

Paid subscription

  • pay-as-you-go per hour of transcription.
  • monthly and annual plans available with additional features.

 

Otter.ai
  • provides real-time transcription for meetings, interviews, lectures, and more.
  • transcribes conversations in real-time, supporting both in-person and virtual meetings.
  • transcribes conversations in real-time, supporting both in-person and virtual meetings.
  • supports transcription in English, French, and Spanish.
  • integrates with platforms like Zoom, Microsoft Teams, and Google Meet.
  • allows sharing and collaboration on transcripts with team members.
  • useful for accurate meeting notes and transcriptions.

 

  • free plan has limitations on transcription minutes and features.

Free version

  • Free Plan: Offers 300 transcription minutes per month.

 

Paid subscription

  • Pro Plan: Provides 1200 minutes per month.
  • Business and Enterprise Plans available for higher usage and team collaboration.

 

 

AI-driven research support

All-round AI tools assisting with a suite of research-related tasks, including writing, discovery, summarization, and data analysis, making the research process more efficient. However, caution is advised when using such tools, as they may sometimes generate content that lacks depth, misinterpret complex concepts, or overlook subtle contextual elements, which could lead to errors or incomplete understanding. Always complement AI-generated outputs with thorough review and expert insight.

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Moxie
  • AI-powered platform for academic research, providing tools for research design, writing assistance, literature reviews, data collection, and publication readiness.
  • Includes integration with premium models like GPT-4 Turbo and Claude 3.5. 
  • useful for research writing and structuring projects.
  • helps refine research questions, literature synthesis, and

 

  • requires critical assessment for accuracy and compliance with academic standards. 

Free version

  • free 7-day trial.

 

Paid subscription

  • full access to over 50 academic tools, including personalized feedback, peer review simulation, and access to professional development sessions

 

 

Coding

AI tools can automate repetitive tasks, improve code quality, and offer real-time feedback, significantly speeding up software development and enhancing understanding for non-experts. However, they may overlook nuances in complex problems, introduce errors, or even lead to unintentional code plagiarism. Careful review and validation are essential to ensure accuracy, originality, and reliability.

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Code Interpreter in GPT-4
  • provides Python code execution to analyze data and solve problems.
  • allows users to describe tasks in natural language, making coding and data analysis more accessible to non-programmers.
  • supports advanced data processing tasks, from regression modeling to creating insightful visualizations like bar charts, scatter plots, and histograms.
  • useful for data analysis, especially for those without coding knowledge. 
  • simplifies learning programming and data science.

 

  • execution is restricted to the sandbox environment—no access to external servers or real-time web searches.
  • code complexity might occasionally exceed processing capabilities.
  • users must ensure sensitive data privacy when using uploaded files.

 

Paid subscription

  • available with ChatGPT Plus. No separate subscription required.

ChatWithGit plugin in GPT-4
  • assists with basic code review and collaboration using GitHub. 
  • useful for coding and collaboration but requires technical knowledge to leverage effectively. 

Free version

  • Free with GitHub integration. No separate paid version.

AskTheCode plugin in GPT-4
  • enabling developers to analyze GitHub repositories and interact with their content directly within the chat interface.
  • answers coding-related questions and troubleshooting. 
  • designed to work with various programming languages, including Python, JavaScript, Rust, and more, ensuring versatility across different codebases.
    Ask The Code Docs
  • useful for coding assistance.
  • developers can streamline their workflow, gaining immediate insights into codebases and enhancing their coding efficiency.

 

  • limited in real-time collaboration capabilities. 
  • effectiveness depends on how you prompt ChatGPT.

Available with GPT-4 model access. 

GitHub Copilot 
  • AI-powered coding assistant.
  • offers autocomplete-style suggestions in supported Integrated Development Environments.
  • useful for streamlining coding workflows.
  • allows developers to ask coding-related questions and receive assistance.
  • reduces repetitive coding tasks 

 

  • limited to well-known languages and programming paradigms.

Free version

  • free trial.

 

Paid subscription

  • various subscription plans (Individual Business, Enterprise) ranging from core to advanced features and customization options.

Amazon Q Developer 
  • AWS-based AI coding assistant for auto-completion and suggestions. 
  • identifies and suggests remediations for vulnerabilities.
  • useful for research projects within the AWS ecosystem.
  • accelerate coding tasks, improve code quality, and maintain security best practices.

 

  • less suited for external environments. 

Free version

  • basic access.

Paid subscription

  • for enterprise use.

Plagiarism checkers

Tool

Main characteristics

Use & limitations

Free | Paid subscription

Turnitin 
  • designed primarily for academic institutions.
  • compares submissions against a vast database of academic papers, student work, and web content.
  • includes AI tools for identifying paraphrased content and contract cheating.
  • suited for institutions to ensure academic integrity in publications.
  • offers deep integration with learning management systems.

 

 

Institutional subscription only

  • Turnitin is used as a tool for plagiarism detection at UHasselt for Bachelor's and Master's theses. Students do not have access through the institutional license.

SafeAssign
  • integrated into Blackboard.
  • compares submissions against a vast database of academic journals, websites, and institutional archives to identify potential instances of plagiarism.
  • generates detailed reports highlighting matched text and providing similarity percentages
  • allows instructors to easily incorporate plagiarism checks into their assessment processes

 

  • may not consistently detect paraphrased content.
  •  currently lacks the capability to detect AI-generated content.

Institutional subscription only

  • At UHasselt, SafeAssign can be used in Blackboard by teaching staff.

Grammarly’s Plagiarism Checker 
  • built-in plagiarism checker alongside grammar and style tools. 
  • checks submissions against ProQuest and online content to identify copied text. 
    .
  • suitable for students, and professionals who need a lightweight, combined plagiarism and grammar checker.

 

  • not as extensive as Turnitin or iThenticate in terms of academic database coverage​.

Free version

  • includes basic grammar checks but no access to the plagiarism checker. 

 

Paid subscription

  • includes full plagiarism detection, grammar, and writing enhancements​.
  • UHasselt students and staff can request a premium account at no personal charge and login with their UHasselt credentials. Registered users can upload documents and use the plagiarism and AI check in the bottom right corner.

iThenticate 
  • built for publishers, academic researchers, and institutions to verify manuscript originality.
  • specializes in research papers, dissertations, and publications​.

 

  • not ideal for student-level checks. 

Paid subscription only

Quetext 
  • compares submissions against the web and academic sources.
  • offers detailed reports with highlighted similarities and citation help.
  • includes an AI-powered 'DeepSearch' feature for in-depth analysis. 
  • user-friendly plagiarism checker.
  • easy to access for individuals​

 

  • offers fewer institutional integrations than Turnitin.

Free version

  • limited plagiarism checks and report details. 

 

Paid subscription

  • unlimited checks and advanced features like DeepSearch.

Unicheck 
  • real-time plagiarism detection by scanning online sources, academic papers, and educational repositories.
  • fast processing and ease of use.
  • integrates smoothly with learning management systems.
  • suitable for schools and educational institutions that need real-time plagiarism detection integrated with LMS platforms.

 

  • not as feature-rich as Turnitin in terms of feedback tools​.

Institutional subscription only

Ithaka S+R generative AI product tracker

Ithaka S+R curates a regularly updated list of Generative AI tools actively used in post-secondary education and research. The list can also be accessed directly as a Google doc. Ithaka S+R is a division of ITHAKA, the U.S. based organization that also operates JSTOR and helps the academic community use digital technologies to improve global access to knowledge and education.

Collection

Support

Practical

Close menu