Studyforge
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle What I Built I built Studyforge , a locally hosted assessment builder for teachers, professors, tutors, and other educators. An educator enters course material or learning objectives, adds details such as the course, grade level, and student instructions, then chooses how many multiple-choice, short-answer,…
This submission for the Hacktoberfest Open-Source AI Challenge Week 2 introduces Studyforge, a locally hosted assessment builder designed for educators such as teachers, professors, and tutors. By inputting course material or learning objectives, educators can customize the assessment with details like course, grade level, and student instructions.
The tool then generates multiple-choice, short-answer, or long-answer questions before providing a draft with an answer key, a browser preview, and a print-ready PDF that can be either a student copy or a teacher copy with the answer key on a separate page. The primary audience includes educators looking to streamline classroom test preparation.
Studyforge operates as a Node.js and Express application, with its browser UI served by the Express server and endpoints available for generating assessments and downloading PDFs. The application leverages Ollama to run the Gemma 3 270M model locally, ensuring that all processing occurs offline, thus eliminating the need for a hosted inference API.
During initial setup, the app downloads the necessary model and dependencies, enabling generation of questions and the main workflow without relying on external hosted services. PDFKit is utilized to create downloadable assessment PDFs.
The advantage of using open-weight models and local inference with Studyforge lies in its independence from closed, hosted APIs. Educators can operate the model on their own machine, maintain control over their course material, and avoid potential fees associated with per-request hosted-model usage. This local setup also enhances adaptability, allowing users to configure different Ollama models to match their hardware specifications or requirements.
However, it's important to note that local inference involves certain trade-offs, including the initial download of the model, dependency on hardware capabilities for generation, and the possibility that smaller models may require thorough review of the content produced.
Currently, Studyforge does not include billing or paid plans. A public deployment or video demo is not available yet, but interested users can run the application locally by following the setup instructions provided in the README. To access the assessment builder, navigate to http://localhost:8000 after completing the setup. The source code for Studyforge is hosted on a GitHub repository, which can be accessed through a specified URL.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
