Dolos
Source code plagiarism detector for programming exercises, with interactive similarity graphs and a free web app.
Upload student submissions, see which ones share solutions and explore the links in your browser.
Dolos is a source code plagiarism detection tool built by Team Dodona at Ghent University for programming courses. It compares submissions to each other and shows who shares solutions, even when the code was edited. It suits teachers and teaching assistants who grade code.
What it is
Dolos analyses a set of source files for similarities between them and presents interactive visualisations such as plagiarism graphs. You can use the hosted web app, run the CLI to open results on a local web server, or call the JavaScript library from your own tools. Results are also exported as CSV files.
Who it's for
It suits instructors running programming assignments who want to compare a class's submissions against each other. It does not check code against the open web or external repositories, and you still judge every flagged pair yourself.
Key features
Free hosted web app
Upload files at dolos.ugent.be and get a report without installing anything.
Interactive similarity graphs
Visualisations show clusters of students sharing solutions and can be used to discuss plagiarism with students.
Many languages via tree-sitter
Parsing relies on the tree-sitter library to support a wide range of programming languages.
Local CLI with web view
dolos run -f web starts a local web server so submissions never leave your machine.
CSV output
Analysis results are available as machine-readable CSV files.
JavaScript library
Embed the detection algorithms in your own Node.js workflow with the dolos-lib package.
Up and running in 3 steps
Our honest verdict
What's great
- Detects shared solutions even when code was modified, using tree-sitter parsing rather than plain text matching.
- A free hosted web app means a teacher can run a check with no install.
- The local CLI and Docker image let you keep student code on your own computer.
- Backed by a peer-reviewed SoftwareX paper from the Dodona team, which you can cite in research.
Where it falls short
- It only compares submissions you give it against each other, so it will not find code copied from sources outside your upload.
- The CLI needs Node.js (release notes for v2.10.0 require Node.js 22 or newer) or Docker, which is a hurdle for non-technical staff.
- The web app is a research-group service, so check your institution's rules before uploading student work to it.
- A high similarity score is a lead for you to review, not proof of misconduct.
Pricing & license
Dolos is MIT-licensed and free to use. The project runs a free web instance at https://dolos.ugent.be and shows no paid plan, so if you self-host the only cost is your own server time.
Open remote AI-training roles
767 open remote AI-training roles, updated November 14, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
Is Dolos free for teachers?
Yes. The README says there is a free instance at dolos.ugent.be and the code is MIT-licensed.
Can I run it without uploading student code?
Yes. Install the CLI with npm or use the Docker image and Dolos serves the results on a local web server.
Which languages does it support?
Many, through the tree-sitter parser library; the Dolos site links to a supported languages list.
