Local AI is having a moment, and we want you to be part of it!

Running through May 24, the Gemma 4 Challenge invites you to explore open models. With the release of Gemma 4, Google's most capable open model family yet, we now have access to native multimodal capabilities, advanced reasoning, a 128K context window, and models that range from running on a Raspberry Pi, to phones, to powering large-scale deployments.

Whether you love to build or love to write, there's a prompt for you and a $3,000 prize pool up for grabs!

Read on to learn more.

Our Prompts

Build With Gemma 4

Your mandate is to build something useful or creative with any Gemma 4 model. The scope is wide open — you can build anything from an IoT integration to a multimodal tool to a long-context reasoning app. What matters is that Gemma 4 is doing real work at the heart of your project.
 

Build With Gemma 4 Submission Template

 

The most compelling submissions will make a clear case for why you chose the model you did and what that model unlocked.

Write About Gemma 4

Your mandate is to publish a post about Gemma 4 that educates, inspires, or sparks curiosity. There's no single right format — what matters is that your post offers something genuine and useful to the community.

Not sure what to write about? Here are some ideas:

  • How-to guide: Walk through setting up and running a Gemma 4 model locally, fine-tuning it for a specific task, or integrating it into a real project
  • Comparison piece: Break down the three Gemma 4 model variants and help readers decide which one is right for their use case
  • Personal essay or opinion piece: Share your experience building with Gemma 4, or make a case for something — what does a model this capable running locally mean for the future of AI?
  • Deep technical breakdown: Explore a specific capability like multimodal input, the 128K context window, or reasoning mode

 

Write About Gemma 4 Submission Template

 

Need inspiration? Check out how Google's own team fine-tuned Gemma 4 with Cloud Run Jobs. That’s the kind of hands-on, shareable knowledge is exactly what we're looking for.

Note: If you are primarily showing off a project, please submit to the Build with Gemma 4 prompt instead! Each submission is only allowed to be eligible for one of the two categories.

Which Model Is Right for You? 🤔

Gemma 4 model family spans three distinct architectures tailored for specific hardware requirements:

  • Small Sizes: 2B and 4B effective parameter models built for ultra-mobile, edge, and browser deployment (e.g., Pixel).
  • Dense: A powerful 31B parameter dense model that bridges the gap between server-grade performance and local execution.
  • Mixture-of-Experts: A highly efficient 26B MoE model designed for high-throughput, advanced reasoning.

No matter which you choose, judges will be looking for intentional model selection — show us why your model was the right tool for the job.

Prizes 🏆

Five winners from the “Build with Gemma 4” prompt will receive:

  • $500 USD cash prize
  • DEV++ Membership
  • Exclusive DEV Badge

Five winners from the “Write About Gemma 4” prompt will receive:

  • $100 USD cash prize
  • DEV++ Membership
  • Exclusive DEV Badge

All participants with a valid submission will receive a completion badge on their DEV profile.

Judging Criteria

Build With Gemma 4 submissions will be evaluated for:

  • Intentional and effective use of the chosen Gemma 4 model
  • Technical implementation and code quality
  • Creativity and originality
  • Usability and user experience

Write About Gemma 4 submissions will be evaluated for:

  • Clarity and depth of explanation
  • Originality of perspective or insight
  • Practical value to the community
  • Quality of writing

Getting Started With Gemma 4 🚀

Gemma 4 is open — you have several ways to get started, including completely free options:

  • Gemini API via Google AI Studio: Access Gemma 4 through the Gemini API.
  • Run locally (free, no credit card required): Download any Gemma 4 model directly from Hugging Face or Kaggle. The E2B model runs on high-end phones — or even a Raspberry Pi 5.
  • OpenRouter (free tier available): Access Gemma 4 31B via OpenRouter's free tier — no credit card required.

Important Dates

  • May 6: Gemma 4 Challenge begins!
  • May 24: Submissions due at 11:59 PM PDT
  • June 4: Winners Announced

We can't wait to see what you build and write! Questions about the challenge? Drop them in the comments below.

Good luck and happy coding!