Gemma

Top Builders

Explore the top contributors showcasing the highest number of app submissions within our community.

Gemma

Gemma is a lightweight, open large language model (LLM) from Google, optimized for efficient AI applications. As part of the Google Gemma family, it uses a transformer-based architecture tailored for responsible and accessible AI usage. Developed as a foundational model, Gemma serves various basic language processing needs, including chatbots, content summarization, and multilingual support.

General
Relese date	February 2024
Author	Google DeepMind in collaboration with Google AI teams
Website	[Google AI Gemma]https://ai.google.dev/gemma
Repository	Google AI Developer Resources
Type	Open-source AI, transformer-based LLM

Key Features

Efficient Deployment: Available in parameter sizes like 2.5B and 7B, Gemma balances capability with efficiency, enabling deployments on both edge devices and cloud infrastructure.
Flexible Tuning Options: Offers pre-trained and instruction-tuned variants, allowing developers to optimize for specific use cases or deploy as-is.
Decoder-Only Transformer Architecture: Uses a streamlined decoder-only design, enabling Gemma 1 to process up to 8192 tokens in one pass for better handling of long-form text.
Safety and Accessibility Tools: Integrates responsible AI features, promoting transparency and safety in AI outputs.

Applications:

Chatbot Development: Optimized for conversational tasks, Gemma provides foundational capabilities for chatbot applications.
Summarization and Paraphrasing: Its pre-trained model structure makes it suitable for summarizing content across languages and contexts.
Multilingual Processing: Supports multilingual inputs, making it adaptable for global applications and translation services.

Get started building with Gemma:

Developers can quickly integrate Gemma into applications by accessing its model weights on Google AI Studio and Kaggle. The model’s lightweight design ensures that it can run efficiently on most hardware configurations, including mobile and edge devices. For optimal performance, utilize frameworks such as Keras or JAX to customize and deploy Gemma for your specific use case. Get started today by exploring the tools and resources available on the Google AI Gemma platform.

Edit on GitHub

Google Gemma AI technology Hackathon projects

Discover innovative solutions crafted with Google Gemma AI technology, developed by our community members during our engaging hackathons.

HUMOS

Robot training is expensive(VLA) and hard. What if a first person view of humans hand can be used for training robots? Why is this helpful? Anyone with a mac or an iphone can then start collect training data. They can get paid for it and robotic data can be accelerated. Here's the IDEA real time camera data is used along with media pipe, sam3 , yolo and vlm so the egocentric data can be enrinched with accurate masks from sam3, reasoning from vlm, mediapipe can detect joints data. All of this is super useful for training robots in a cheap and fast way. From lack of data , abundance of data is achieved fast. This is more useful especially for specialized tasks, where only certain humans can do and they are in remote places Object tracking with persistent IDs across frames Zero-shot state classification via SigLIP — 200x faster than VLM for open/closed/ajar labels Navigation state classification via VLM (doors, drawers, handles → open/closed/ajar/blocked) Temporal diff — VLM compares consecutive frames to detect state transitions Navigation timeline — per-object state timeline with colored bars and transition events Hand-object interactions via MediaPipe Ground truth export — structured JSON with per-frame annotations Accuracy evaluation — compare predictions against manual labels Live perception — real-time webcam inference with auto-recording and post-analysis H.264 video export — browser-playable annotated videos with in-app preview Per-frame timing — inference latency breakdown per model stage Tried Gemini 3 Flash, Cosmos and gemma for vlm

Based Codey

Vibe coding alone wasn’t enough. I wanted to become a real, hands-on coder without losing creativity or momentum. I was frustrated by tools that assume perfect conditions: a laptop, stable internet, long focus windows, and formal education paths. Real life doesn’t work like that. So I built Based Codey to let me learn, build, and ship from anywhere, even from my phone. The deeper inspiration was democratizing access to real software development—so learning to code isn’t locked behind privilege, hardware, or gatekeeping. What it does Based Codey is an AI-powered, production-oriented coding environment that allows the user to build real applications end-to-end in a conversational, hands-on way, with the option for education. It combines: – Various AI agents with differing reasoning levels (from snippet repair to project level reasoning) – A live editor and preview environment – Tasking, annotations, and iterative workflows – Mobile-first usability so I can ship from anywhere – It’s not just “generate code”—it’s learning by building, with real files, real structure, and real ownership. – Still, if you want just "generate code" it can do that.

CareSight AI - Autonomous Healthcare Assistant

CareSight AI is an autonomous, multimodal healthcare assistant designed to reduce clinical paperwork and improve care readiness across the patient journey. The system uses computer vision, OCR, speech recognition, and LLM reasoning (Gemma 3N) to perceive and structure unorganized healthcare data such as documents, images, and spoken responses. Patients begin with a mobile web app to upload IDs, insurance cards, lab reports, and clinical images, or provide voice responses. CareSight AI extracts structured data, validates completeness, and prepares an intake summary. During nurse check-in, the system assists by auto-filling forms, flagging inconsistencies, and supporting preliminary screenings such as cognitive memory tests by recording sessions, transcribing responses, and automatically scoring recall accuracy. Doctors receive a concise, explainable summary highlighting key findings and uncertainties. CareSight AI functions as an autonomous agent following a perception–reasoning–action loop, supporting clinicians without replacing them and enabling scalable, remote-ready healthcare workflows.

RoboGripAI

This project presents a simulation-first robotic system designed to perform structured physical tasks through reliable interaction with objects and its environment. The system focuses on practical task execution rather than complex physics modeling, ensuring repeatability, robustness, and measurable performance across varied simulated conditions. Simulation-first robotic system performing structured physical tasks such as pick-and-place, sorting, and simple assembly. Designed for repeatable execution under varied conditions, with basic failure handling, environmental interaction, and measurable performance metrics. A key emphasis of the system is reliability under dynamic conditions. The simulation introduces variations such as object position changes, minor environmental disturbances, and task sequence modifications. The robot is designed to adapt to these variations while maintaining consistent task success rates. Basic failure handling mechanisms are implemented, including reattempt strategies for failed grasps, collision avoidance corrections, and task state recovery protocols. The framework incorporates structured task sequencing and state-based control logic to ensure deterministic and repeatable behavior. Performance is evaluated using clear metrics such as task completion rate, execution time, grasp accuracy, recovery success rate, and system stability across multiple trials. The modular system design allows scalability for additional tasks or integration with advanced planning algorithms. By prioritizing repeatability, robustness, and measurable outcomes, this solution demonstrates practical robotic task automation in a controlled simulated environment, aligning with real-world industrial and research use cases. Overall, the project showcases a dependable robotic manipulation framework that bridges perception, decision-making, and action in a simulation-first setting, delivering consistent and benchmark-driven task execution.

FreshCredit Arc UCP Integration

FreshCredit Arc UCP Integration demonstrates a next-generation commerce and credit infrastructure where users retain control over their financial data while intelligent agents handle discovery, verification, and payment execution. The project combines Circle’s Arc USDC payment rails with Google’s Universal Commerce Protocol (UCP) to enable agent-based interactions between consumers and providers. Users authenticate via decentralized and enterprise-grade identity systems, maintain portable, user-owned financial reports at the edge, and authorize transactions only when value is exchanged. Instead of monetizing report access, FreshCredit aligns incentives by charging only on completed transactions. UCP enables standardized offer discovery and intent matching, while Arc enables instant USDC settlement across wallets and networks. Together, they form a privacy-preserving, interoperable alternative to legacy credit bureaus, optimized for AI agents, automation, and global commerce.

Awebo AI

Awebo.AI is your AI product manager, built on Google's Gemini 2, designed to navigate the communication of small to medium primarily-remote companies. We get that things can get lost in translation between teams, and that's where we come in. We're already live on Slack. We have also fine-tuned Gemini 2 model to understand the nuances of your conversations and provide real-time support for product development. Think of it as having a super-smart product manager in your pocket, always ready to offer insights, generate ideas, and keep everyone in the loop. Right now, we're hyped about our upcoming Linear integration, which will streamline your workflow and help you manage projects with insane efficiency. But that's just the beginning. We're constantly pushing the boundaries of what's possible with AI, and we're working on a bunch of other integrations to make Awebo.AI an indispensable part of your product development process. Stay tuned, because things are about to get really interesting.