Open Source Hack Day · Hacktoberfest

Event Format

This event is an in-person Hacktoberfest Hack Day. Participants will work in teams to build an original project using open-source or open-weight AI technology and submit their work through MLH OrganizerHQ Challenges.

Challenge 1 — Best Open-Source AI Project

• Every team must build an original project in which open-source or open-weight AI is an important part of how the project works.
• The project must be published in a public GitHub repository.
• The repository must clearly state an open-source license such as MIT, Apache-2.0, GPL-3.0, or another OSI-approved license.
• Teams must submit their project through OrganizerHQ Challenges before the announced submission deadline.
• Teams should clearly explain the AI technology, model, dependencies, and contribution in their README and demonstration.
• One winning team will be selected for this challenge.
• Every member of the winning team receives a DEV Badge; additional MLH+DEV prize swag may be provided depending on availability and event onboarding.

Challenge 2 — Best Use of Gemma 4

• This is the assigned Hack Days partner prize category for ITM Open Source Hack Day 2026.
• The project must make meaningful use of Gemma 4 through the Gemini API.
• Teams should identify the Gemma model used in their README and demonstrate the integration in their code or live demonstration.
• Gemma should contribute meaningfully to the project's experience or functionality, rather than being included only as a superficial feature.
• The project must be submitted through OrganizerHQ and must satisfy the partner challenge requirements.
• The winning team will receive the prize specified in the event's official MLH onboarding materials. Prize availability may vary by region.

Project Development Rules

• Project-specific work should begin at the time announced by the organizers, following the MLH Standard Hackathon Rules unless the organizers clearly announce a different approach before project work begins.
• Participants may use open-source libraries, APIs, datasets, frameworks, and other permitted development tools.
• Teams are responsible for complying with the licenses and terms of all third-party software, models, APIs, and datasets they use.
• Teams must clearly identify major technologies and AI models used in their project.
• AI coding assistants and other development tools may be used, provided the resulting work is understood and responsibly submitted by the team.

Submission Rules

• Every project being considered for a prize must be submitted through OrganizerHQ Challenges.
• One team member should submit the project on behalf of the team.
• Submissions must include the project link, project name, description, technologies used, and any available demo URL.
• Teams must select every challenge for which their project is eligible, including Best Open-Source AI Project and Best Use of Gemma 4 where applicable.
• Late submissions will not be considered unless the organizers officially extend the submission deadline for all teams.

Judging Criteria

• Technical Implementation: quality, functionality, and technical depth of the project.
• Use of AI: meaningful and effective use of open-source/open-weight AI and, for the Gemma category, meaningful use of Gemma 4.
• Originality: creativity, innovation, and thoughtfulness of the solution.
• Open Source: public repository, clear licensing, reproducibility, and quality of documentation.
• Demo & Explanation: clarity of the demonstration and the team's ability to explain what they built and how it works.
• Impact & Usefulness: how effectively the project addresses a real problem or creates a useful experience.

General Rules

• Participants must register and check in physically at the venue through OrganizerHQ.
• The event is free to attend and is conducted entirely in person.
• Participants must follow the MLH Code of Conduct and all reasonable instructions from event organizers and venue staff.
• Plagiarism, misrepresentation of another person's work, manipulation of submissions, malicious activity, or other unfair means may result in disqualification.
• Teams must respect software, model, dataset, API, and content licenses and terms of use.