Welcome to the May 2025 edition of the Global Agent Hackathon!
Join us for a month-long, open-source AI Agent Hackathon — open to all builders and dreamers working on agents, RAG, tool use, and multi-agent systems.
Powered by Agno, Exa, Firecrawl, Mem0, Unwind AI, Browser Use, mcp.run, Graphlit, Lutra AI, Potpie and more.
Win over $25,000 in cash and credits for your favorite AI tools for the most ambitious Agent projects!
Best Use of Agno (see details):
Projects showcasing the best use case of Agno will get following cash prizes:
- 🏆 GRAND PRIZE: $5,000 🏆
- 🥇 1 winner: $2,000
- 🥈 5 winners: $1,000 each
- 🥉 10 winners: $500 each
- 🏅 10 winners: $300 each
The $300, $500 and $1,000 prizes will be awarded on a rolling basis. The $2,000 and $5,000 prizes will be awarded at the end of the hackathon.
Judges: Dirk Brand, Yash Solanky, Monali Dambre
Best use of Browser Use (see details):
Projects showcasing the best use case of Browser Use will get following cash prizes:
- 🏆 GRAND PRIZE: $2,500 🏆
- 🥉 2 winners: $500 each
- 🥈 10 winners: $150 each
The $150 prizes will be awarded on a rolling basis. The $500 and $2,500 prizes will be awarded at the end of the hackathon.
Judges: Larsen Cundric, Mert Ünsal and Aitor Mato
Best use of Potpie 🥧 (see details):
Projects showcasing the best use case of Potpie will get following cash prizes:
- 🏆 GRAND PRIZE: $200 🏆
- 🥉 3 winners: $100 each
- 🥈 5 winners: 3 months of Potpie Pro FREE!
All the prizes will be awarded at the end of the hackathon.
Judges: Aditi
- Startup plan for free for a whole year to the best project built using Browserbase
Judges: Lindsay Gilson
Every participant gets free $50 Exa API credits!
- To get the credits use the coupon code
EXA50APIby heading to the Exa API Dashboard.
- Every one is entitled to 5k Firecrawl credits, enter the coupon code
hackathon5kon the Firecrawl Dashboard
- Mem0 is offering pro plan for free for 3 months to the best project built using Mem0. (details coming soon)
Judges: Taranjeet Singh
- MCP.Run is offering a free 3 month pro account to any participant, using
AGNOHACK2025at checkout. Thank you MCP.Run for sponsoring the hackathon!
- Please see details
Groq is the lowest cost inference provider with no tradeoffs on speed and they're offering $10 in Groq API credits to any participant (millions of tokens for all available models, including Llama 4 Scout and Maverick) and $300 in Groq API credits to the best project built using Groq.
- Fill out this form for $10 in Groq API credits
- Best use of Groq will get $300 in Groq API credits + feature on Groq Applications Showcase
- The top 5 projects will be featured in the trending Awesome LLM Apps repo, giving your work massive visibility in the AI community.
Judges: Shubham Saboo
Resources for each sponsor are available in the resources folder.
- April 15th - Hackathon starts
- June 1st - Hackathon ends
- Submit your project by creating a GitHub issue in this repo following the provided template.
- Get category assignment
- The Hackathon team will review your submission and assign it to a prize category.
- You can update your issue as you build, and we'll update the prize category accordingly.
- Prize categories can be updated until the hackathon ends.
- Categories include Best use of Agno, Firecrawl, Mem0, Graphlit and Browser Use.
- Build your project and submit a PR for your project in the submissions folder - we'll review, judge, and provide feedback. Note: you can start building as we review your submission.
- When your PR is approved, you join the pool to win the assigned cash prize and/or credits.
- Checkout the How to Win section for more details.
- Winning projects will be moved to the winners folder.
Your submission will be judged on:
- Completed PR with working code in the Hackathon GitHub repo.
- A judge will review the PR and provide feedback.
- Once approved, the PR will be merged and the project will be eligible for the prize pool.
- Demo Video (2–3 mins) showing your agent in action.
- Comprehensive README.md including:
- Overview of the idea
- Explanation of how it works
- Technologies/tools used
- All setup instructions (API keys, signups, etc.)
- Code Quality: Error-free execution and passing basic review.
- Completeness: Core functionality implemented as described in your original submission.
Feel free to share your progress on X or LinkedIn with the hashtag #GlobalAgentHackathon — even small updates can get you noticed and help build community momentum. This is completely optional, but always welcome and appreciated!
Include the following in your submission:
Project Title: A clear, concise, and engaging title for your project.
Overview of the Idea: Briefly describe your project in simple terms. What problem or opportunity does it address?
Project Goal: What do you hope to achieve or demonstrate with this project?
How It Works: Provide a step-by-step explanation or high-level breakdown of your Agentic System and how users will interact with it:
- User Flow: Clearly describe the experience from the user's perspective
- Core Functionality: List the key features your agent will offer
- Multimodal Elements (if applicable): Specify if you're using text, images, audio, video, or combinations
Tools Used: List what you're planning to use, including any new tools you need to create.
UI Approach: Describe the user interface for your project.
Visuals: Add mockups, diagrams, or sketches that demonstrate your idea or flow.
Team Information:
- Team Lead: GitHub handle and role
- Team Members: GitHub handles and roles
- Background/Experience: Brief team background
If you're an AI company and want to help shape the next generation of Agent builders - we'd love to work with you!
Reach out to support@agno.com to get involved.
Let's make this the largest open-source AI Agent hackathon ever!
See you in the arena.