AMD Developer Hackathon: ACT II
Build AI agents and high-performance AI apps on AMD GPUs in the cloud I Prize Pool $15,000
To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account.
ADP members can access AI tutorials, experts, and community support. In addition, new members can claim a free month of DeepLearning.ai pro, $100 in AMD GPU cloud credits, and $50 in Fireworks AI API credits.
All hackathon participants will receive on hackathon launch additional compute and API credits for their projects.
Ready to build what's next on AMD? Sign up with AMD below to secure your spot.
The AMD Developer Hackathon: ACT II is a hands-on event for developers, founders, engineers, and builders who want to push what's possible with AI on real infrastructure.
At the center of it: AI Agents - a space to explore intelligent workflows, automation, and real AI applications. Whether you're just starting out or already building, you can jump in and start creating.
You'll be working with AMD AI Developer Cloud, ROCm, and Fireworks AI API, all fully in the cloud, so you can focus on building instead of setup.
All AMD Developer Program Participants Get
AMD AI Developer Program members unlock cloud credits, training resources, expert access, and opportunities to get their work recognized.
☁️ $100 AMD Developer Cloud GPU Credits
New AMD AI Developer Program members receive $100 in cloud credits redeemable on AMD Developer Cloud with powerful AMD GPUs to start building.
🔥 $50 Fireworks AI API Credits
New AMD AI Developer Program members receive $50 in Fireworks AI API credits to access models hosted on AMD-hardware to start building right away.
🎓 AMD AI Academy
Access AMD AI Academy courses, technical resources, tutorials, documentation, and developer learning content.
🧑💻 AMD Expert Access
Connect with AMD engineers and technical experts through member community channels and office hours.
💬 Private Community Access
Get access to community channels, personalized updates, and early registration opportunities for AMD developer events.
🏆 Project Recognition
Get opportunities for project recognition, including potential showcase on AMD social channels or at AMD developer events.
🚀 Fully Cloud-Based
Everything runs fully in the cloud, so no local hardware setup is required to participate and build.
AMD GPU Cloud Access and Credits
On hackathon start, details for AMD GPU cloud access and credits will be provided. Benefits are provided through the AMD AI Developer Program and may be subject to AMD's terms, eligibility, and availability.
Refer friends and earn points
Invite other builders with your personal referral link. Once you're approved, you can generate your link from the event dashboard and start sharing.
Get approved
Register for the hackathon and wait until your enrollment is approved.
Generate your link
Open your dashboard and click "Generate referral link."
Share and track
Share your link and track referral activity from your dashboard, live page, or profile.
How points work
Earn 200 points when someone joins through your link and submits a project. Referred participants can earn 100 points when they submit.
Minimum threshold to qualify for the referral prizes
You must refer at least 100 approved participants to be eligible for referral prizes — even if you finish in the top 3 on the leaderboard.
Referral activity appears after invited participants enroll and get approved.
Three tracks. Real AMD hardware. Pick your challenge.
Whether you're shipping your first AI agent or building your next startup, ACT II has a track for you. Build on AMD Developer Cloud, ROCm, and Fireworks AI API credits.
All submissions must be containerized.
Hybrid Token-Efficient Routing Agent
Build an AI agent that gets the job done using the least tokens possible.
Tasks are revealed at kickoff. Your agent must complete each one autonomously by deciding in real time whether to use a local model or call a remote model via Fireworks AI credits. The goal: pick the cheapest option every time, without falling below the accuracy threshold.
Every submission is scored on a standardized environment. You can develop and test on any hardware, but final scoring runs on this standardized environment only. Local models must therefore be sized to run within these constraints, so routing intelligence wins, not raw compute power.
We recommend running a local eval step to check your output quality before submitting.
Want to fine-tune your router? Go for it. Prompt-based and fine-tuned approaches are scored exactly the same way: token count and output accuracy.
Models to be used will be revealed on launch day.
Video Captioning
Four styles. Endless creativity. Let your humor bone shine.
You'll receive a fixed set of short video clips. Your pipeline must generate a caption or summary for each one in four distinct styles: formal, sarcastic, humorous-tech, and humorous-non-tech.
Models are accessed via Fireworks AI API credits. Fine-tuning is explicitly permitted — you may also train your own captioner and use it alongside or instead of prompting. Use open datasets, build your own, or both. Models to be used will be revealed on launch day.
Unicorn Track
Your idea. AMD infrastructure. No benchmarks, no constraints — just build.
Use any open-source models and frameworks alongside AMD GPUs and/or Fireworks AI API credits to build a product- or startup-oriented project.
There is no fixed performance benchmark. Submissions are not scored on speed, token usage, or accuracy. Judges are looking for creativity, originality, completeness, use of AMD platforms, and product/market potential. Think startup pitch, not benchmark run.
Technology & Access
Build using AMD Developer Cloud with AMD GPUs, ROCm, Fireworks AI API, and open-source AI frameworks.
Everything runs fully in the cloud, so teams can focus on building, experimenting, benchmarking, and scaling AI workloads without managing infrastructure.
AMD Developer Cloud
On-demand access to AMD GPUs for training, fine-tuning, benchmarking, and deploying AI workloads.
Fireworks AI API
Fast, scalable API access to AMD-hardware models for inference, fine-tuning, and building AI-powered pipelines.
ROCm
ROCm is AMD's open-source GPU computing platform for AI, ML, and high-performance workloads on AMD GPUs.
AMD Training and Learning Resources
Participants also get access to curated AI learning resources, tutorials, hands-on labs, and training materials to help them build faster.
Technology Partner: Gemma
Gemma, Google's family of lightweight open models, is joining AMD Developer Hackathon: ACT II. Use Gemma in your project for a chance to win the Best Use of Gemma Models challenge.
Gemma is Google DeepMind's family of lightweight, open-weight models built from the same research as Gemini, available to participants through AMD Developer Cloud and Fireworks AI.
- • Add Gemma to your routing, captioning, or agent workflows
- • Use Gemma's multimodal and reasoning capabilities within your track's requirements
- • Check your track's model restrictions before building, as access varies by track
Access
Via Fireworks AI credits
Compute
AMD Developer Cloud
License
Apache 2.0
Sign-up
Through AMD AI Developer Program
For this hackathon, Gemma is accessed through Fireworks AI and AMD Developer Cloud. Usage can be drawn from your hackathon credits, or from the $50 Fireworks AI credits available to new AMD AI Developer Program members. No separate sign-up is required.
- Sign up for the AMD AI Developer Program to access AMD Developer Cloud and Fireworks AI credits.
- Use your Fireworks AI credits to call a Gemma model directly through the Fireworks API.
- Run your application on AMD Developer Cloud using your allocated GPU access.
- Check your track's requirements before choosing a model, as model restrictions vary by track.
Prizes
🏆 Total Prize Pool: $15,000
What to submit?
Submit your project through the lablab.ai platform before the deadline. Make sure all fields below are completed.
📋 Basic Information
- • Project Title
- • Short Description
- • Long Description
- • Technology and Category Tags
📸 Cover Image and Presentation
- • Cover Image
- • Video Presentation
- • Slide Presentation
💻 App Hosting and Code Repository
- • Public GitHub Repository
- • Demo Application Platform
- • Application URL
Important Requirements
• Your project must be submitted through the lablab.ai platform before the deadline.
• All submissions must be containerized.
• Your GitHub repository must be public and include a README with setup and usage instructions.
• The application must be runnable using the provided instructions.
⏰ Submission Deadline
Check the Event Schedule tab — the deadline is always up to date and shown in your local timezone.
For general lablab.ai submission guidance, please visit Submission Guidelines.
lablab.ai — Hackathon Submissions Process
Judging Criteria
Scoring varies by track. Tracks 1 and 2 are ranked via leaderboard. Track 3 — Unicorn Track is evaluated by judges using the criteria below.
Creativity and Originality
The uniqueness and creativity of the solution, highlighting novel approaches and new behaviors.
Product/Market Potential
The startup or product vision — how compelling and viable the idea is in a real market context.
Completeness
How fully realized and functional the submitted project is.
Use of AMD Platforms
How meaningfully AMD infrastructure is incorporated into the project.
Hackathon Details
Everything you need to know before you register and start building.
🛠️ How to participate
To access credits and be approved for the hackathon, participants must sign up for the AMD AI Developer Program.
The hackathon takes place online on the lablab.ai platform. Click the Enroll button on this page to register.
🦸 Who can join?
Anyone with a passion for AI is welcome. To participate, simply sign up on lablab.ai and register for the AMD AI Developer Program.
👥 What about teams?
If you don't have a team yet, connect with other builders and find teammates on the lablab.ai Discord Server. Join the conversation and start building with others from day one.
💬 AMD Discord Server
Join the AMD Discord for infrastructure questions, hardware, software, and AMD AI Academy tutorials and support throughout the hackathon.
Note: To access hackathon credits and resources, make sure you have signed up for the AMD AI Developer Program before the hackathon starts.
Speakers, Mentors & Judges
Event Schedule
- To be announced
Teams: AMD Developer Hackathon: ACT II
Check out the roster and find teams to join
.png&w=3840&q=75)
.png&w=3840&q=75)
