About

About Grongy

Grongy is an academic research and writing assistant for students and researchers. It searches the scholarly indexes, more than 320 million records in OpenAlex among them, turns a topic and a research question into an editable chapter plan, and suggests drafts section by section from the sources in your library. Its interface is in English, Turkish, German, Spanish and French.

  • 1,000+researchers and students
  • 320M+scholarly works (OpenAlex, October 2026)
  • 10,000+journal and university styles from the CSL repository (Plus and Pro)

Why it exists

The product came out of one ordinary failure. You ask a chatbot for a source, it hands you a reference with the right authors, a plausible title and a real-looking journal, and the paper may not exist. You find out at the worst possible moment — in a supervisor’s comments, or at your defense.

So the design starts from the opposite end. A draft can cite only publications in your library; when it cites a paper it found in a search, that paper is added to your library. When the sources do not support a sentence, the sentence stays uncited instead of receiving an invented reference. That produces a weaker-looking first draft and a better thing to hand in, and that trade is the whole product.

What it will not do

These are not aspirations. Some sit in the code, some in the rules the model is given, some in the terms of use.

  • No feature for getting past plagiarism software or AI detectors, and the terms of use prohibit using it that way.
  • No invented sources: a citation in generated text with no match in your library is dropped from the text. Not inventing data, participants, measurements or quotations is the model’s rule; checking them against their sources is yours.
  • A draft is never presented as ready to submit. What comes back is raw material to check, cut and rewrite.
  • No invented ratings or review counts published to win stars in search results.
  • No student surveillance sold to institutions: there is no institutional login, and the librarian report holds only aggregate counts.

How it works, briefly

The short version of what happens between a research question and a chapter.

Sources from the indexes, not a model’s memory

It searches OpenAlex, Crossref and DOAJ; PubMed and arXiv when your topic fits them; Turkish indexes for Turkish searches; Google Books and Open Library for books. Where a legal open-access PDF exists, it finds it, through Unpaywall among others.

Citations bound to your library

Drafting and autocomplete write from the publications in your library and from candidates found in a search; a candidate it cites is added to your library. A citation with no match in your library does not enter the generated text. Chat can also search the web.

A document in its submission format

Headings, numbering, a table of contents, captions, footnotes and page setup; Word, PDF, LaTeX or Markdown export; nine built-in citation styles and more than 10,000 CSL styles for the reference list (the automatic reference list is on Plus and Pro).

Your work starts in your browser

Documents live in your browser and back up to your own account’s private storage when you sign in. Where it is stored and which services are involved is in our privacy policy.

Who is behind it

Grongy is built by a small independent team. For anything about the product, the fastest route is email — a person reads it.

Contact

Questions, university enquiries, bug reports and press: [email protected].