Launch IO
$25K in a Month, And That's Only Chapter One
- Prize Pool$25,000+
- Duration17 Jun - 21 Jul
- LocationOnline
30 Days to $25K and a Whole Lot More.
Overview
Whether you're an indie hacker, AI tinkerer, or dev curious about autonomous agents, the Launch IO Hackathon is your chance to build with cutting-edge AI infrastructure. Create practical, real-world tools using IO Intelligence's AI Models API or AI Agents API, and take the first step in shaping how AI agents operate in the wild. From finance agents to crypto copilots and smart research bots, we want to see what you can launch when creativity meets capability
Are you ready to build the next great AI agent?
Eligibility
- Open to All:
Builders, developers, tinkerers — if you're excited about agentic AI, you're welcome.
- Beginner-Friendly:
Both newcomers and experienced devs can join. Prior IO experience is not required.
- Flexible Teams:
Solo builders or teams of up to 5 participants are allowed.
- Build Rules:
You can start from scratch or build on something existing — just clearly show what was created during the hackathon.
- Tech Stack:
Your project must use at least one IO Intelligence tool: AI Models API and/or AI Agent API
Tracks
Competitive Track: Autonomous Agents in the Real World
Build a smart, useful AI agent that performs real-world tasks with minimal human input. Whether it's a crypto co-pilot, research bot, content summarizer, or productivity tool — your project should show how IO Intelligence can power intelligent, self-running tools that solve real problems.
What to Explore:
- API Integration & Action Execution:
Agents that read from and write to live systems — triggering workflows, sending messages, updating records, or manipulating hardware via APIs.
- Sensor-Aware Decision Making::
Agents that consume real-world data from cameras, IoT devices, or microphones to make context-sensitive decisions.
- Identity & Access Management:
Agents that act securely and responsibly within enterprise systems — respecting permission boundaries and audit requirements.
- Human-AI Interaction:
Agents that collaborate with humans in shared environments — providing status updates, asking for input, or acting on delegated authority.
Requirements:
- Use the AI Models API and/or AI Agent API
- Submit working code, a short demo video, and documentation
- Projects will be judged on creativity, functionality, usefulness, and presentation
Beginner Track: Agent Framework Editor
New to IO or agentic AI? Start here. This guided track helps you learn how the Agent Workflow editor works and rewards you for sharing your ideas.
What You'll Do:
- Explore and familiarize yourself with the IO Agent Workflow Editor
- Record and submit a short video (2-5 mins) explaining how it works
- Share your ideas built on Agent Workflow Editor.
- Submissions will be judged on presentation and clarity, understanding, creativity and presentation
Prizes and Bounties
Placement
Amount
Hackathon Flow:
Registration
17 Jun - 14 Jul
Build & Submit
1 Jul - 23 Jul
Review
24 Jul - 31 Jul
Demo Day (Judging & Awards)
1 Aug
FAQ:
Do I need prior experience with IO Intelligence to join?
Nope! The hackathon is beginner-friendly. We'll provide docs, templates, and Discord support to help you get started.
Can I join alone or do I need a team?
You can join solo or in a team of up to 5 people. If you need a team, we'll have a space on Discord to connect with others.
What can I build in this hackathon?
Anything that uses IO's AI Models API and/or AI Agent API. Some examples: research assistants, crypto copilots, content tools, daily task agents.
Is there any cost to participate?
No — the hackathon is completely free to join.
What do I need to submit?
You'll need a GitHub repo, a short demo video (2-5 mins), and a brief write-up describing your project and how it uses IO.
Can I use other tools or libraries in my project?
Yes! You can use any open-source tools or external APIs, as long as your project includes IO Intelligence.
Rules:
- This hackathon is open to solo or team participation of not more than 5 members.
- You can either invite your friends to form a team or you can request other teams to add you as a member.
- The hackathon will start on July 1, 2025 12:00 PM UTC and end at Jul 14, 2025 11:59 PM UTC, you work on your hack during the allotted time only.
- It's an online hackathon, you can participate from anywhere.
- Once the hackathon starts, you will get an option to submit your hack, you can submit as many times as you want, the last hack will be considered as the final submission.
- Your hack must include the use of at least 1 product in the IO Intelligence suite (AI Models API and/or AI Agent API). You may use open source libraries and other freely available systems / services such as Google Maps, Facebook Connect, Twitter feeds etc.
- The intellectual property of your code belongs only to your team.
Resource Center
Quick answers to questions you may have. Can't find what you're looking for? Check out our documentation
Judging Criteria - Competitive Track
Creativity
Is the idea unique or does it use an interesting approach?
Usefulness
Does it solve a real problem or offer strong utility?
Functionality
Does it run smoothly and achieve its intended goal?
Presentation
Is the project clearly explained in the video and write-up?
Completeness
Is the idea fully built or just a partial concept?
Demo Quality
Does the demo effectively show what the agent does?
Code & Documentation
Is the code well-structured, and is setup clearly explained?
Beginner Track
Clarity
How clearly you explain the Agent Framework
Understanding
How well you understand the core ideas
Creativity
Is your proposed agent idea thoughtful or original?
Presentation
Is your video simple, engaging, and easy to follow?
Winners
🥇 1st Place - IO.AIRGPT
What we love:
IO.AIRGPT impressed judges with its originality and depth. It combined multiple IO components to create a general-purpose AI interface with contextual memory and agentic behavior. The project showed strong potential as a developer tool, with smooth functionality and thoughtful presentation.
🥈 2nd Place - Sniffle
What we love:
Sniffle is an autonomous agent that scans Solana's memecoin ecosystem to detect early signs of trending tokens. It stood out for its sharp focus on real-time on-chain analysis and timely alerts, giving traders an edge in spotting microcap surges. Judges praised its functionality and strong execution in a high-risk, high-reward niche.
🥉 3rd Place - ZeroTouch: Supply Chain AI
What we love:
ZeroTouch created an autonomous agent that monitors real-world disruptions (like port delays or strikes) and responds by rerouting supply chains. Its strong use of live data, real-world APIs, and autonomous decision-making made it one of the most ambitious and useful entries in the hackathon.
🎖️ Honourable Mentions
Novel
A multi-agent startup assistant that automates research, planning, and document generation for founders. It stood out for its polished execution, strong agent orchestration, and seamless use of IO's APIs.
PulseLensAgent
A workplace productivity tool that used biometric signals to trigger contextual task switching and alerts — showing strong creativity and integration.
Subnet Scout
A network visualization and monitoring agent for decentralized compute nodes. It made backend complexity feel approachable and useful.
Meeting Co-Pilot
An AI agent that joins meetings, transcribes, and summarizes action items. It was a great example of human-AI collaboration and practical utility.
JadenX.AI
A character-driven agent platform with creative interaction models and strong storytelling potential. While still early, the ambition and design caught our eye.
