MOSS vs Dolos vs JPlag, compared
MOSS, Dolos and JPlag are the three free tools most programming courses use. All three compare a batch of submissions with each other by structure, and none of them checks the web or detects AI. They differ in how you run them, which languages they read, and how clear the report is. Here is how they compare, and where a paid checker fits.
Checks every file against classmates, the web and AI, in 65+ languages.
At a glance
| MOSS | JPlag | Dolos | Codequiry | |
|---|---|---|---|---|
| From | Stanford, 1994 | KIT, 1996 | Ghent University | Codequiry |
| Cost | Free for academic use | Free, open source | Free, open source | Paid |
| How you run it | Perl script to a hosted server | Java app on your machine | Online, or on your machine | Web app, LMS, API |
| Compares submissions with each other | Yes | Yes | Yes | Yes |
| Leaves out starter code | Yes | Yes | Yes | Yes |
| Checks the web | No | No | No | Yes |
| Detects AI-written code | No | No | No | Yes |
| Languages | About 25 | About 20 | About 20 | 65+ |
| Report | Web pages, kept 14 days | Local report viewer | Interactive graphs and side-by-side | Dashboard, exportable, kept in your account |
Language counts change as the projects add support. Check each project's own documentation for the current list.
MOSS
MOSS (Measure of Software Similarity) was written by Alex Aiken at Stanford and has been the default in CS courses for decades. You request an account by email, receive a Perl script, and send each batch from a terminal. Results come back as linked web pages.
- Good at: a quick peer comparison once it is set up. Its winnowing fingerprints ignore names, comments and formatting.
- Watch out for: results are deleted after 14 days, there is no support, and restructured copies can score low. We show that on a worked example in What MOSS misses.
JPlag
JPlag began at the Karlsruhe Institute of Technology and is now an open-source project on GitHub. You download it and run it with Java on your own computer, so submissions never leave your machine. It compares token sequences with an algorithm called greedy string tiling, and it opens results in a report viewer in your browser.
- Good at: private, offline checking, and a well-maintained codebase you can inspect.
- Watch out for: you need Java and a command line, and you manage the files and the reports yourself.
Dolos
Dolos comes from Ghent University. It parses code with tree-sitter and compares fingerprints, and it can run online or on your own machine. Its report is the most visual of the three: a graph of which students are connected, clusters, and side-by-side views.
- Good at: getting started quickly and seeing the shape of a class at a glance.
- Watch out for: like the others, it only compares the files you give it.
Need more than peer comparison?
Check the same batch against past terms, the web and AI.
Where Codequiry fits
All three free tools answer one question well: did students in this batch share code? Most copying now comes from somewhere else, and that is the gap Codequiry fills:
- The web. GitHub, Stack Overflow, Software Heritage and the indexed web, with a link to each source.
- Past terms. Earlier submissions to the course, without rebuilding the batch by hand.
- AI-written code. A detector for ChatGPT, Claude, Gemini and Copilot output, with its reasons shown.
- Your LMS and your records. Canvas, Moodle and others, and every report kept in your account.
It is a paid product. For a one-off, a free check compares files with each other, the same job the free tools do, with no account.
How to choose
- You want free and simple: start with Dolos.
- Code must stay on your machine: use JPlag.
- You already have a MOSS account and a workflow: keep MOSS, and save each report before it expires.
- You need the web, past terms or AI checked: none of the free tools does it. Use Codequiry, alone or next to one of them.
More detail on each: MOSS alternative, how to detect code plagiarism.
Frequently asked questions
What is the difference between MOSS, JPlag and Dolos?
All three compare a batch of submissions with each other by structure. MOSS is a free hosted service from Stanford that you use through a script. JPlag is open-source software from KIT that you run yourself with Java. Dolos is open source from Ghent University, with an online version and a visual report.
Which is the best free code plagiarism checker?
It depends on how you work. Dolos is the easiest to start with and has the clearest reports. JPlag suits people who want to run everything on their own machine. MOSS is the long-standing default and is simple once registered.
Do MOSS, JPlag or Dolos check the web?
No. All three compare only the files you give them. Code copied from GitHub or Stack Overflow is caught only if another submission in the batch copied the same thing.
Can MOSS, JPlag or Dolos detect AI-generated code?
No. None of them has an AI detector. Codequiry checks for AI-written code alongside peer and web checks.
Is Dolos better than MOSS?
For many instructors, yes: it is open source, it can run online or on your own machine, and its reports are easier to read. Both compare submissions with each other only.
Is JPlag free?
Yes. JPlag is open source and free to use. You download it and run it with Java on your own computer.
Why pay for Codequiry if the others are free?
For what the free tools cannot do: checking the web and earlier terms, detecting AI-written code, LMS integration, and keeping every report. If peer comparison alone is enough for your course, a free tool may be all you need.
Check what the free tools cannot see
Classmates, past terms, the web and AI in one check, with the lines side by side.
Get Started Just one assignment? Run a single check