Vibe Code with Gemini 3 Pro – Hackathon Report (Dec 2025)
Vibe Code with Gemini 3 Pro – Hackathon Report (Dec 2025)
What is Vibe Code with Gemini 3 Pro
On 5 December 2025 Google DeepMind launched a hackathon called “Vibe Code with Gemini 3 Pro in AI Studio.” The event runs via the public competition platform Kaggle Competition page on Kaggle.
The aim is to let developers worldwide build AI-powered applications using Gemini 3 Pro and Google AI Studio. The challenge encourages building working prototypes — combining reasoning, code, and multimodal inputs such as text, images or video — not just data-science notebooks or simple experiments AI Studio – Vibe Code page.
Official Hackathon Details
The publicly available information lists the following:
| Field | Detail |
|---|---|
| Start date | 5 December 2025 Kaggle overview |
| Submission deadline | 12 December 2025 Kaggle overview |
| Prize pool | US $500,000 in credits for Gemini API usage Competition rules |
| Submission requirements | Projects must use Gemini 3 Pro in AI Studio; external tools, datasets or APIs allowed if core logic runs through Gemini 3 Pro Rules page |
| Allowed domains / themes | Science, health, education, accessibility, business, technology or other fields per announcement Public announcement on LinkedIn |
The official Kaggle page is already live and open for registration and submissions Competition page.
What is Gemini 3 Pro – Core Features
Gemini 3 Pro is the model that powers this hackathon. Launched by Google DeepMind in November 2025, it brings advanced reasoning, multimodal understanding, coding, and agentic abilities Gemini 3 announcement blog (blog.google).
Features:
- Gemini 3 Pro supports multimodal input and output: text, image, video, code, and mixed media Gemini 3 Pro model page (cloud.google.com).
- The model combines strong reasoning with code generation and media handling, enabling full-app prototypes instead of simple chat output Gemini 3 announcement blog (blog.google).
Because of these capabilities the hackathon invites building complex, real-world applications — prototypes that mix reasoning, data, media, code, and user interaction.
Public Launch and Announcement
The hackathon launch was publicly announced by Google DeepMind via social and professional channels. A post on LinkedIn announced:
“Vibe Code with Gemini 3 Pro in AI Studio hosted by Google DeepMind, build your best app … Prize Pool: $500,000 in Credits … Hackathon Timeline: Dec 5–12, 2025” LinkedIn announcement (linkedin.com).
On X (formerly Twitter), a post from the official competition handle flagged the start of the hackathon and referenced the prize and Gemini 3 Pro model X post about hackathon start (x.com).
That confirms the hackathon is real, organised by Google DeepMind, with infrastructure provided by Kaggle.
Early Public Activity – Projects, Notebooks, Community Signals
Even though the hackathon is ongoing, there are already public signs of activity — shared notebooks, early builds, and community-driven idea lists — showing what people try to build with Gemini 3 Pro + AI Studio.
Shared Notebooks and Early Builds
On the competition’s code/notebook section there are publicly visible notebooks referencing gemini-3-pro-preview, which indicates early experimentation or prototype builds (possibly intended for hackathon submission) Kaggle code submissions tab (kaggle.com).
These public artifacts illustrate how developers are using the model for real work — not just quick tests — combining reasoning, data processing, maybe UI or media handling.
External Coverage and Project Idea Lists
A recent article summarises five early public “vibe coding” projects built with Gemini 3. These range from interactive 3D screens to sign-language tutors, webcam portals, AI-powered podcast hosts and panorama / image-editing tools Analytics Vidhya: Gemini 3 vibe coding projects (analyticsvidhya.com).
Some projects reportedly include 3D interactive particle screens controlled by hand gestures, real-time video-enabled sign-language learning apps with confidence scoring, and multimedia podcast host tools. While these come from public community reports and not necessarily the hackathon submission list, they illustrate the creative potential people are exploring with Gemini 3 + AI Studio.
What You Can Do Now – If You Want to Participate or Explore
If you read this and want to join or experiment:
- Go to the official Kaggle competition page and sign up Competition page.
- Open Google AI Studio and make sure to select Gemini 3 Pro or the preview model AI Studio Vibe Code.
- Choose a project idea that fits reasoning, multimodal input / output or code generation — for example document processing, interactive media apps, data analysis tools, or custom utilities.
- Build a minimal viable prototype quickly — focus on functionality and demoability, rather than polish.
- Submit your project through Kaggle with the required deliverables (code/notebook, demo link/video) under the rules.
Because the prize is given as API credits, even participants who don’t “win” may benefit: they gain access to a powerful model and can continue development or experimentation without initial infrastructure costs.
Why Gemini 3 Pro and This Hackathon Matter in 2025
Gemini 3 Pro reflects a shift in AI: moving beyond text-only models to models that combine reasoning, media, coding, and tool use. For developers and creators this opens new possibilities — building AI-powered apps, interactive tools, media pipelines, or hybrid software + AI systems.
With Vibe Code the barrier to experiment lowers. Hosted on a public competition platform and open to a broad audience, the hackathon encourages rapid prototyping.
Many third-party platforms and tools already start integrating Gemini 3 Pro — for example developer environments or enterprise tools seeking to leverage its reasoning and code generation strengths Cloud blog: Gemini 3 for enterprise (cloud.google.com).
This hackathon acts as a public testbed — for developers, makers, students, or anyone curious — to try building real apps with next-generation models quickly. It offers a rare chance to experiment at scale with minimal entry cost and broad visibility.
Public Resources & Useful Links
- Vibe Code with Gemini 3 Pro – official Kaggle page
- Gemini 3 Pro model announcement by Google DeepMind
- Gemini 3 Pro official model page
- X post from Kaggle announcing hackathon
- LinkedIn announcement by Google DeepMind / Kaggle
- Article summarising early public Gemini 3 vibe-coding projects
JSON-LD Schema
{
"@context": "https://schema.org",
"@type": "NewsArticle",
"headline": "Vibe Code with Gemini 3 Pro – Hackathon Report (Dec 2025)",
"datePublished": "2025-12-07",
"author": {
"@type": "Organization",
"name": "VibeCoding Team"
},
"description": "Report on Vibe Code with Gemini 3 Pro in AI Studio hackathon by Google DeepMind on Kaggle — includes details, model info, and early public activity.",
"mainEntityOfPage": {
"@type": "WebPage",
"@id": "https://vibecoding.app/gemini-3-pro-kaggle-hackathon-report"
},
"publisher": {
"@type": "Organization",
"name": "VibeCoding",
"sameAs": "https://vibecoding.app"
}
}
About VibeCoding Team
VibeCoding Team is part of the Vibe Coding team, passionate about helping developers discover and master the tools that make coding more productive, enjoyable, and impactful. From AI assistants to productivity frameworks, we curate and review the best development resources to keep you at the forefront of software engineering innovation.
