Articles
AI hackathons: the complete guide
What an AI hackathon is, who runs them, what teams build, what you can win, and how to take part on Starfield Labs from enrolling to submitting.
By Starfield LabsPublished

An AI hackathon is a time-boxed event where people form teams and build a working application around AI models. Most run for a weekend or a few days, though some are a single day and some open challenges run for weeks. This guide covers the whole field: what makes an AI hackathon different, who runs them, what teams build, what is on offer, and how taking part works on Starfield Labs.
If this is your first event, read What to expect at your first AI hackathon for the hour-by-hour feel and a packing list. If you do not code, start with AI hackathons for beginners with no coding experience. If you want to know how projects are scored, read How hackathon judges score projects.
What makes an AI hackathon different?
A general hackathon asks for working software. An AI hackathon asks for working software where a model does real work: reading, writing, reasoning, seeing, listening or acting. A few things follow from that.
- You build on models, you rarely train them. Teams call a hosted model through an API, or run an open-weight model on their own machine. Training a model from scratch does not fit in a weekend.
- A working demo counts for more than a big plan. A small loop that runs on real input tends to beat an ambitious idea that only exists on slides.
- Mixed teams do well. A developer who can call a model API, someone who knows the problem first hand, someone who can design the screens, and someone who can tell the story in a short video make a strong team.
- The model's limits are part of the work. Models can be wrong with confidence. Teams that test on real examples and say honestly where their project fails tend to be easier to trust.
Who runs AI hackathons?
AI hackathons come from several places, each with its own flavour:
- Hackathon platforms such as Starfield Labs, which run online events open to anyone, often with an AI company or another organization as a partner that sets the challenge.
- AI companies and tool makers, which run events to get developers building on their models or APIs.
- Universities and student societies, often over a weekend, sometimes with a theme from a research group.
- Companies, which run internal hackathons for their own staff.
- Conferences, governments and non-profits, often with a social-impact theme.
The format differs, but the core stays the same: a challenge, a deadline, a submission and a judged result.
What do teams build?
Most AI hackathon projects fall into a handful of fields. A few examples per field:
| Field | Example projects |
|---|---|
| Productivity | A meeting summarizer, an email drafting assistant, a code review helper |
| Education | A tutoring bot, a language practice partner, a quiz generator |
| Health and wellness | A note structurer for clinicians, a symptom timeline for patients, a mood journal |
| Creative tools | A script outliner, a storyboard maker, a podcast show-notes writer |
| Developer tools | A documentation bot, a test generator, a debugging assistant |
| Social good | An accessibility helper, a plain-language form explainer, a community alert system |
For more than 50 ideas sorted by field, with scope cuts for a weekend, see AI hackathon project ideas you can demo in a weekend.
What can you win?
Each event sets its own prizes. Across the field you will see:
- Cash prizes, the most common kind.
- Credits or tools from a partner, such as usage of their models or platform.
- Follow-up support, such as time with a partner's team or an invitation to keep building.
- Recognition: a public project page and something to show on your profile or in a job search.
On Starfield Labs, each event's page lists its prizes. Starfield's operations team contacts the winning teams about their prizes. The prize is rarely the only win. A finished project you can show, and the people you built it with, often last longer.
How do you take part on Starfield Labs?
Starfield Labs runs its hackathons online, and taking part is free. Here is the path from finding an event to submitting.
- Sign in. You can sign in with a code sent to your email, or with GitHub or Google. You need to be at least 16, the minimum age in the Terms; an event may ask for 18 or more in its own rules.
- Find a hackathon. Browse the list at starfieldlabs.ai/ai-hackathons. Each event's page holds its challenge, timeline, rules and prizes.
- Enroll. Press Enroll on the event's page. Some events ask staff to approve each enrollment first, and Starfield Labs shows you when yours is waiting.
- Form a team. Every member of a team enrolls on their own. You can create a team, open to anyone or by invite only, up to the event's maximum size. You can also join one: ask to join an open team in the team finder, accept a team leader's invite, or open a team link a leader shares. Working alone is fine too, since a team of one counts as a team.
- Build. Your team has its own chat on Starfield Labs. Announcements from the organizers reach you on Starfield Labs, by email and on Discord, and the Starfield Labs Discord is also where people meet, ask questions and find teammates. If you build with an AI coding agent, Starfield Labs has an MCP connection, so your agent can read the event's challenge too.
- Submit. Your team's project form walks through up to six steps: Team, Project, Presentation, Video, Links and Review. A title and a short summary are always needed. The rest, such as a cover image, a video upload, a repository or a demo link, depends on what the event's form asks. Review shows what is still missing, and Submit opens once everything required is filled. You can keep editing a submitted project until the deadline.
- Judging and results. Judges score each project against the event's criteria, and no judge scores their own team's project. Once judging is done, the results are published on Starfield Labs.
Staff of Starfield Labs cannot take part in hackathons. For more help, see the Help desk.
How long does an AI hackathon last?
Most run from one to three days of building. A 24-hour event leaves room for one model step and one screen. A 48 to 72-hour event leaves room for a second step, such as search over your own data or a tool the model can call. Longer challenges, which run for weeks, usually expect a more finished product. The timeline on the event's page is the one that counts.
What should a submission hold?
What the event's form asks, and nothing it does not. Across events, a strong submission usually has:
- A clear title and a one-paragraph summary of the problem and the user
- A short demo video that shows the project working on real input
- A repository with a README that says how to run it
- A demo link judges can open, if the event asks for one
- An honest note on what does not work yet
For a time plan that gets you there, see How to win an AI hackathon.
Can you use existing code and libraries?
Usually, yes, for libraries, frameworks, starter templates and public models. Most events expect the project itself, the part that makes it yours, to be built during the event. Each event's rules say where its line sits, so read them before you start, and say in your write-up what you brought with you.
Where to go next
- Pick an API and a stack: Best AI APIs for hackathons
- Plan your 72 hours: How to win an AI hackathon
- Pick an idea: AI hackathon project ideas you can demo in a weekend
- Learn a tool first: the tutorials at starfieldlabs.ai/ai-tutorials
Then find your event at starfieldlabs.ai/ai-hackathons and enroll.
Keep reading
Related articles and tutorials
- Tutorial

PaLM2 Tutorial: Create your own AI-powered Virtual Assistant
AI hackathon tutorial: build a chat-style virtual assistant with Google's PaLM2 API and a Streamlit app in Python, step by step.
- Article

Best AI APIs for hackathons
How to choose a model API for a hackathon project, what Claude, GPT, Gemini, IBM Granite and open-weight models are each good at, and how to keep your stack small. Facts checked 2026-11-19.
- Article

How to win an AI hackathon
Build the right team, pick a problem you can demo, aim at the event's criteria, and follow a 72-hour plan from locking the idea to submitting early.
- Tutorial

OpenAI Whisper tutorial: How to use Whisper to transcribe a YouTube video
AI hackathon tutorial: download YouTube audio with Pytube and transcribe it with OpenAI Whisper in a few lines of Python.