On-site

ExecuTorch Hackathon

Name: ExecuTorch Hackathon
Start: 2026-06-27T20:00:00.000Z
End: 2026-06-28T20:00:00.000Z
Location: lablab.ai

Build on-device AI with ExecuTorch on Snapdragon

🌉 On-site Only Hackathon | 📅 June 27–28, 2026 | 📍 San Francisco, CA

This is an in-person hackathon in San Francisco for selected participants. Capacity is limited to 30 teams, teams must include 3–5 members, and there is no online or hybrid participation.

Build real-time AI apps that run directly on Snapdragon-powered mobile devices using ExecuTorch. Selected teams will get Samsung Galaxy S25 Ultra devices on-site for building, testing, and live demos.

📍 On-site Venue (June 27–28):

The Web Data Loft by Bright Data

625 2nd St, San Francisco, CA

📱 Samsung Galaxy S25 Ultra devices provided on-site for selected teams

🧠 Learn from Qualcomm and Meta experts through workshops, mentorship, and hands-on support

🏆 Prizes include Meta Quest 3 headsets, Ray-Ban Meta AI Glasses, and post-event project support

🧑‍💻 Apply now to build the future of on-device AI with ExecuTorch and Snapdragon.

Once pre-approved, you will receive further instructions on how to complete the acceptance process.

Application deadline: June 17, 2026, at 11:59 PM PDT.

About the Hackathon

Build AI applications that run on-device

The ExecuTorch Hackathon is a two-day, on-site builder event in San Francisco focused on taking AI applications from PyTorch to ExecuTorch and running them on Snapdragon-powered Samsung Galaxy S25 Ultra devices.

Register interest first, then complete the selection process

This is a proposal-based, on-site only hackathon with limited capacity. Registering your interest does not guarantee attendance. Applications will be reviewed based on a combination of your Edge AI idea, background, team status, and ability to attend on-site in San Francisco.

Format On-site only

Capacity 30 teams

Teams 3–5 members

Apply by June 17

How the selection process works

This hackathon works a little differently — and that's what makes it exciting. There are two phases to reach the final stage in San Francisco.

Phase 1

Idea Contest

Open to everyone

1
Apply and create your team on Lablab. Teams of 3 to 5. The stronger the team, the stronger the idea. No team or idea yet? Apply anyway — there'll be plenty of chances to find teammates and shape an idea along the way.
2
Add your project idea to your team description by June 17. After registering, create your team on Lablab and describe your Edge AI concept in the team description: what you're building, why it matters, and what problem it solves. We'll review all submissions and pick the top 30 to move forward.
3
Need teammates? Flag it in the screening form and we'll help you connect. We also run matchmaking sessions during this phase (updates go out via newsletter). Join our Discord and check #team-matchmaking-executorch-hackathon and #brainstorming-ideas.
4
Ideas will be reviewed after June 17, 2026. We will review all submissions together with the Qualcomm team. The best 30 ideas and their team will be invited to join the hackathon in San Francisco.

Phase 2

The Hackathon in San Francisco

On-site final

1
Selected teams get an official invitation and move to the final stage - the in-person hackathon in San Francisco, with details about the pre-hack workshop and next steps.
2
Final team matchmaking. After selection, we run one more round to make sure all teams are complete and strong. Need more members? Looking to join a selected team? We'll help make that connection. .
3
Not selected? Teams whose ideas were not chosen will be notified and will not be enrolled in the hackathon on Lablab.

Trademark notice: Snapdragon and Qualcomm branded products are products of Qualcomm Technologies, Inc. and/or its subsidiaries.

What to Expect On-site

Two days of building, testing, and demoing on real mobile hardware

This is an on-site only hackathon experience in San Francisco. Selected teams will build together, get technical support, work with Snapdragon-powered mobile devices, and demo their applications live.

✅ Check-in & Device Access

Teams will check in on-site and receive access to the devices and tools needed to build and test their applications.

🎤 Kick-off & Masterclass

The event begins with sponsor introductions and a technical overview of PyTorch, ExecuTorch, and recommended build tools.

🛠 Hands-on Building

Teams will spend the hackathon building, optimizing, and testing AI applications that run locally on mobile devices.

🧑‍🏫 Mentor Support

Industry experts will be available on-site to help teams stay unblocked, make technical decisions, and keep moving.

📱 Live App Demos

Each team will demo their application and highlight the edge AI technologies, performance, and user experience behind it.

🏆 Judging & Awards

Projects will be evaluated by judges, with an additional Team’s Choice Award selected through participant voting.

Device loaner agreement: Teams using loaned devices must complete a loaner agreement before receiving hardware. Devices are collected before final judging.

Plus: food, swag, networking, and celebration

Expect meals during the event, participant swag, time to connect with other builders and experts, and a closing social hour after winners are announced.

The Challenge

Build a real Edge AI application

Your mission is to propose, build, and demo a practical AI application that uses PyTorch and ExecuTorch to run locally on Snapdragon-powered hardware.

Main Challenge

Mobile On-Device AI with ExecuTorch on Snapdragon

Create a responsive, user-facing application that demonstrates why Edge AI matters. The strongest projects will show a clear reason for local execution, whether through lower latency, offline capability, privacy-sensitive processing, energy efficiency, or a smoother real-time mobile experience.

📱 User-facing application

Build something people could realistically use on a mobile device, not only a backend model or isolated technical demo.

⚙️ Edge-ready execution

Show how your application can run locally, use device-side capabilities, and remain practical for real-time interaction.

🎤 Live demo direction

Your project should have a clear path toward a working on-site demo that can be explained and presented in 5 minutes.

Suggested project directions

• On-device vision: real-time detection, recognition, scene understanding, accessibility, or camera-based tools.
• Multimodal assistants: mobile assistants that combine text, images, audio, or contextual input.
• Generative AI on mobile: local generation, summarization, transformation, or creative user-facing workflows.
• Privacy-first AI: applications where keeping data on-device creates clear value.
• Real-time edge intelligence: fast, low-latency tools for everyday mobile use cases.

What a strong proposal shows

• A clear Edge AI use case suited for on-device inference.
• A realistic scope that can be built during the hackathon timeframe.
• A clear reason for using PyTorch, ExecuTorch, and Snapdragon.
• A specific user problem, workflow, or experience the application improves.
• A path toward a working demo that can be presented live on-site.

Proposal note: Each team should submit one Edge AI use case proposal. Proposal review is part of the application and shortlisting process.

Technology Partners & Resources

Use these resources to prepare, build, test, and get support during the ExecuTorch Hackathon.

Host Technology

Qualcomm + Snapdragon

Developer resources, AI Hub tools, sample apps, and Qualcomm support for building and optimizing on Snapdragon-powered hardware.

On-device AI Runtime

Meta + ExecuTorch

ExecuTorch helps teams move PyTorch models toward efficient on-device execution for mobile AI applications.

Build Support

GitHub Copilot

Teams may use GitHub Copilot as an AI pair programmer during development and may be considered for the Copilot-Powered Build Award selected by GitHub judges.

Hardware Partner

Samsung Galaxy S25 Ultra

Shortlisted teams will build and test on Snapdragon-powered Samsung Galaxy S25 Ultra devices, subject to onsite availability and loaner agreement requirements.

Developer Resources

Start with the core Qualcomm and AI Hub resources, then use the sample apps and support channels during the build.

Third-Party Tools

Neo4J Graph Builder

Previous & Participant-Provided Apps

The following resources are shared to help participants explore examples and starting points fairly.

Voice Stress Detection Model Tutor.AI Sample App R.E.D.A.C.T. Sample App GitHub Copilot

Need support?

Mentors will be available on-site, and participants can also use the support communities below.

Join Qualcomm Discord Join lablab.ai Discord Join AI Hub Slack

Awards & Recognition

🏆 Prizes

Standout teams will be recognized through judge evaluation, participant voting, and a GitHub-selected Copilot-powered build award.

Selected by Judges

🥇

Top Award

Awarded to the strongest overall project based on the official judging criteria.

Meta Quest 3 512GB All-in-One Mixed-Reality Headset for each member of the winning team
Qualcomm DevRel support to help complete the application and prepare it for mobile deployment.
Blog and live stream opportunities to showcase the project after the hackathon

Participant Vote

🗳️

Team’s Choice Award

Awarded to the team selected by participant popular vote.

Ray-Ban Meta AI Glasses for each member of the winning team
Qualcomm DevRel support to help complete the application and prepare it for mobile deployment.
Blog and live stream opportunities to showcase the project after the hackathon

Selected by GitHub Judges

💻

Copilot-Powered Build Award

A special recognition for teams that use GitHub Copilot creatively and effectively during the build process.

Recognition Creative and effective use of GitHub Copilot

Potential Rewards Copilot Pro+ access and GitHub for Startups credits, subject to eligibility

Note: The Top Award winners cannot also win the Team’s Choice Award. Prize eligibility may be subject to sponsor terms, event rules, availability, and final review.

Please note: Participation in lablab.ai hackathons is voluntary. Prizes and opportunities depend on eligibility, availability, and third-party sponsors, for whom lablab.ai is not responsible. Hackathon rules, prizes, and terms may change or be canceled at our discretion. Submissions must be original and MIT-compliant. Prize distribution may take up to 90 days.

Full details: Voluntary Participation & Prize Terms

What to submit?

This hackathon includes a proposal-based application process before the event and a final project submission during the on-site hackathon.

📋 Basic Information

• Project Title
• Short Description
• Long Description
• Technology & Category Tags

📸 Cover Image and Presentation

• Cover Image
• Video Presentation
• Slide Presentation

💻 App Hosting & Code Repository

• Public GitHub Repository
• Demo Application Platform
• Application URL

Important Requirements

To be considered for participation and prizes, teams must follow the proposal and final submission requirements below.

Before the hackathon: Team Proposal

• Each team must submit one proposal.
• The proposal must describe an Edge AI use case application.
• The application should use open-source software.
• The application should run natively on a Snapdragon-powered mobile device using PyTorch / ExecuTorch.
• The proposal must be the work and/or idea solely owned by the team members.

During the hackathon: Final Project

• The application must be provided in a public GitHub repository.
• The repository must include a README with app description, team names/emails, setup instructions, run/usage instructions, and an open-source license.
• The application must be runnable using the provided instructions.
• The application and most components must run locally on device. Hybrid edge/cloud is acceptable, but the majority should run locally.
• The final GitHub repository must be submitted through the official Microsoft Form shared by the sponsor at the beginning of the on-site event.

Recommended but optional

• Tests and testing instructions to verify the app setup.
• Notes section with additional information not covered in the application description.
• References used while developing the application.
• Well-commented code.

For general lablab.ai submission guidance, please visit Submission Guidelines.

Judging Criteria

Final submissions will be evaluated using the official weighted criteria provided for the ExecuTorch Hackathon.

40%

Technical Implementation

Evaluation based on NPU utilization, latency and performance, and energy efficiency.

25%

Application Use Case & Innovation

Evaluation through the lens of problem solving, creativity and uniqueness, and user experience.

15%

Local Processing & Privacy

Evaluation based on on-device execution and privacy and security.

10%

Deployment & Accessibility

Evaluation based on ease of installation and use.

10%

Presentation & Documentation

Evaluation based on the clarity of explanation during the presentation, and code quality and documentation.

The Top Award is selected by judges. The Team’s Choice Award is selected by participant popular vote, and teams cannot vote for their own project.

Event Information

Hackathon Details

A few practical details to help you prepare before applying and joining the on-site hackathon experience.

📍 Where and when

The hackathon will take place on-site in San Francisco on June 27–28, 2026. Selected teams will build, test, and demo their applications in person.

👥 Team participation

Teams must include 3 to 5 members. If you do not already have a full team, you can still apply and use the lablab.ai Discord to connect with other builders.

Join lablab.ai Discord

🛠️ How to participate

Submit the application form by June 17, 2026. This is a proposal-based application, and shortlisted teams will receive further instructions on how to complete the acceptance process.

📚 Helpful guides

Review the general hackathon guidance and getting started resources before the event.

Hackathon Guidelines Getting Started Guide

🧠 Prepare with lablab.ai

Explore the lablab.ai AI Tech pages to learn about available technologies, and visit the tutorials page for practical examples, guides, and inspiration before the hackathon.

Reminder: Please apply only if you can attend on-site in San Francisco and participate as part of a 3–5 member team.