Omnara is hosting a one-day hackathon on 10/24 at https://primitive.space/. Teams of up to four will build a product that uses agents and deploy it by the end of the day. Solo entries are welcome. Prizes for the top three teams are listed below, and more prizes and partners will be announced this week.

What to build

You can build anything with agents, as long as it's deployed and people can use it. Here are a few examples to show the range.

  • An agentic MCP server that audits a website for SEO, load times, and how easily agents can use it
  • A code review bot you can talk to in GitHub pull requests
  • A generative UI, where the model builds the interface as you use it
  • An always-on assistant connected to your phone, computer, or home devices
  • A game you play through chat, like a mystery you solve by questioning AI suspects
  • A dashboard that sends research agents out in the background and collects what they find
  • A service that watches something, like prices or a competitor's site, and acts when it changes

Deployment

The bar is higher than a demo. Every project has to be deployed and usable by someone outside your team by 5:30 pm. If it stops working when you close your laptop, it doesn't count. Judging puts a lot of weight on whether your project could launch as a standalone product today.

Tools and credits

You can use any tools, models, and providers you want. We're also giving every attendee credits from Omnara and our partners that cover a full stack. With them you can host on Freestyle or Unikraft, use Supabase for your database and auth, run your agents on Omnara, give them web access with Exa and context.dev, send and receive email with Primitive, and take payments with Stripe.

  • $500 in Omnara model credits (OpenAI, Anthropic, and open-source models)
  • $25 in Supabase credits (a free month of Pro)
  • $250 in Stripe credits and free Stripe Atlas incorporation
  • $234 in Unikraft sandbox credits (six free months of the Team plan)
  • $100 in Freestyle sandbox credits
  • $50 in context.dev credits

Schedule

Food and drinks are provided throughout the day.

  • 10:00 am — Doors open, breakfast
  • 10:30 am — Kickoff
  • 11:00 am — Start building
  • 1:00 pm — Lunch
  • 5:30 pm — Submissions close
  • 6:00 pm — Top demos
  • 6:45 pm — Awards
  • 7:00 pm — Drinks
  • 8:00 pm — End

Requirements

What to build

Build a product that uses agents and deploy it so people outside your team can use it. It can be a web app, an MCP server, a bot, an API, or anything else, as long as it keeps running when your laptop is closed.

What to submit

Submit on Devpost by 5:30 pm. You can keep editing until then. Each submission needs:

  • A live link to your project, added under "Try it out." It should work from the judges' own devices and stay up through awards at 6:45 pm. If it isn't a website, explain how to reach it, and if it needs an account, include a test login or let judges sign up.
  • A short description of what it does, who it's for, how it uses agents, and what judges should try first.

Hackathon Sponsors

Prizes

3 non-cash prizes
First Place
1 winner

​$5k Omnara model credits, ​$5k Supabase credits, $3k Stripe credits, $5k Exa credits, $6k Freestyle sandbox credits, $5k context.dev credits

Second Place
1 winner

$3k Omnara model credits, $3k Supabase credits, $2k Stripe credits, $3k Exa credits, $3k Freestyle sandbox credits, $3k context.dev credits

Third Place
1 winner

$1k Omnara model credits, ​$1k Supabase credits, $1k Stripe credits, $1k Exa credits, ​$1k Freestyle sandbox credits, $1k context.dev credits

Devpost Achievements

Submitting to this hackathon could earn you:

Judges

Ishaan Sehgal

Ishaan Sehgal
Omnara

Flor Ronsman De Vry

Flor Ronsman De Vry
Alchemy

Pratyush Saini

Pratyush Saini
ByteAsk

Mayank Pahadia

Mayank Pahadia
Tinder

Ayaan Gazali

Ayaan Gazali

Leon Yao

Leon Yao
CTO / Whitespace

Alex Tung

Alex Tung
CEO / Whitespace

Asher Dale

Asher Dale
Omnara

Christian Sparks

Christian Sparks
Omnara

Ben Swerdlow

Ben Swerdlow
Freestyle

Judging Criteria

  • Production-ready
    Could this launch as a standalone product today? Someone outside your team can start using it without help, and it works end to end and holds up while they use it.
  • Agentic use
    How much work do the agents do? We look at whether they take actions, run multi-step tasks, work in the background, or act on their own, and how well that fits into the product.
  • Creativity
    How good is it overall? The originality of the idea, the quality of the execution, and the things that are hard to put a number on.

Questions? Email the hackathon manager

Invite others to compete

Hackathon sponsors

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.