NYC  SF  Online        Jobs   Deals  
    Sign in  
 
 
NYC Tech
Events Weekly Newsletter!
*
 
COMING UP

SF Open Source AI Week
Oct 18 - Oct 26

Newark Tech Week
Oct 12 - Oct 20

NY App Week
Oct 19 - Oct 22
 
 
 
 
 
 
 
 
 
 
 
[ SF Open Source AI Week ]
Gary Event 
$60K in cash prizes + $5K in hardware awards.
Oct 24 (Sat) @ 09:00 AM       FREE
Venue, To Be Announced, SF

 
      Register      
 
$60,000 in Cash Prizes + $5000+ Hardware Awards

Build the Future of Edge & Physical AI.

Join us at the legendary AGI House for the builder finale to Open Source AI Week.

Built from the same world-class research & technology as Gemini, Gemma 4 is the most capable open model family you can run on your hardware. We invite you to invent the future of embodied intelligence & edge applications with Gemma. Spend an intensive, high-speed day building 100% local applications running on bare metal. Get hands-on experience with the latest Gemma models & push the absolute limits of edge silicon.

The Gemma family runs across the whole hardware spectrum, from sub-1B parameter models that fit on a single-board computer to 31B models on a robot with a local GPU. Which model you reach for, & what you run it on, is up to you. The only rule is that it runs on the device in front of you, with no cloud anywhere in the loop.

We're convening 100 curated builders in Hillsborough, working in teams of three to four alongside the Google DeepMind Gemma team & hardware partners. Bring your laptop, your phone, & your robot if you have one.

$60,000 in Cash Prizes + $5000+ Hardware Awards

Two tracks. Real hardware. An entirely on-device challenge. Build with Gemma for a share of $45,000 in cash prizes, plus hardware awards from our partners.

Best in Show: End-to-End Android & Pixel AI Experience - $15,000 Cash (+ Google Pixel 11 Hardware): Awarded for the best end-to-end on-device AI experience built on Android & Google Pixel.
Best Use of a Gemma Model
Track 1: Mobile AI (Best use of a Gemma model in a Mobile App) - $15,000 Total:
1st Place: $7,500 cash + Exclusive Gemma Break Glass Trophy
2nd Place: $5,000 cash
3rd Place: $2,500 cash
Track 2: Physical AI, IoT & Robotics (Best use of a Gemma model in IoT / Robotics / Embedded) - $15,000 Total:
1st Place: $7,500 cash + Exclusive Gemma Break Glass Trophy
2nd Place: $5,000 cash
3rd Place: $2,500 cash
Best Use of LiteRT (Google AI Edge) - $15,000 Total:
1st Place: $7,500 cash
2nd Place: $5,000 cash
3rd Place: $2,500 cash
Additional Awards
Best Use of LiteRT $15,000
Sponsored by Google AI Edge.

Best Use of Pixel / TPU - 2 Pixel 11
Best Use of Raspberry Pi - 5 Raspberry Pi 5
Best Use of Coralboard - 5 Coral Dev Boards
Qualcomm Hardware Awards
Qualcomm is also providing hardware prize packages for two winning teams, with prizes for every team member.

First Place $1,350 in prizes per team member

A Prusa CORE One+ 3D printer
An Arduino VENTUNO Q, Modulinos, & a webcam to keep
Second Place $450 in prizes per team member

An iFixit Soldering Tool Kit
An Arduino UNO Q, Modulinos, a webcam, & a USB OTG Hub to keep
Second-place winners can take their UNO Q boards, Modulinos, webcams, & USB OTG hubs home on the day. All first-place prizes & the iFixit Soldering Tool Kits will be shipped to winners after the event.

Build On-Device with LiteRT

Bring your Gemma-powered ideas to life with LiteRT, Google's framework for high-performance on-device AI.

Use LiteRT-LM to run Gemma locally, & explore LiteRT's tools for model optimization & hardware acceleration on supported devices. Whether you're building a mobile assistant or an embedded application, LiteRT helps you turn AI models into experiences that run directly on your hardware.

What you might build

Think about what a ship's computer or a tricorder actually does: a device that sees, hears, reasons & acts on its own, with nothing to phone home to. A wearable that narrates the world to someone who can't see it. A drone that reads a room & decides for itself. A robot that learns a chore from watching you once. If it feels like science fiction, that's the right direction.

How the day works

Every project must use a Gemma-based model & run entirely on-device. Judges rotate through the workbenches science-fair style, & the top six teams demo on stage before awards.

Who should apply

Engineers & researchers in embedded systems, robotics, mobile & on-device ML, & founders building for the edge. We're looking for people who have already built with real hardware & want to show us what they can do with Gemma on it.

Space is limited. Apply to attend.
 
 
 
 
About    Feedback    Press    Terms    Gary's Red Tie
 
© 2026 GarysGuide