December 2025
Introducing Gemma Scope 2
Helping the AI safety community deepen understanding of complex language model behavior
Slide 1 of 3
Gemma
A collection of lightweight, state-of-the-art open models built from the same technology that powers our Gemini models*
Gemma 3
A family of lightweight models with multimodal understanding and unparalleled multilingual capabilities for more intelligent applications.
Gemma 3n
Our powerful and efficient open model designed to run locally on phones, tablets, and laptops.
Our most advanced open models help developers create AI applications that run wherever users need them — from workstations to laptops and even phones.
Slide 1 of 9
December 2025
Introducing FunctionGemma
Bringing bespoke function calling to the edge
December 2025
Introducing T5Gemma 2
The next generation of encoder-decoder models
September 2025
Introducing VaultGemma
The world's most capable differentially private LLM
September 2025
Introducing EmbeddingGemma
The best-in-class open model for on-device embeddings
August 2025
Introducing Gemma 3 270M
The compact model for hyper-efficient AI
July 2025
Introducing T5Gemma
A new collection of encoder-decoder Gemma models
July 2025
Introducing MedGemma
Our most capable open models for health AI development
May 2025
Introducing Gemma 3n
Models
Slide 1 of 13
Gemma 3
Multimodal capabilities and wide language support in sizes from 270M to 27B parameters
Gemma 3n
Mobile-first architecture optimized for low-latency audio and visual understanding
MedGemma
Optimized for medical text and image comprehension
VaultGemma
Trained from the ground up with differential privacy
EmbeddingGemma
Generates high-quality text embeddings efficiently
DolphinGemma
Helps scientists study how dolphins communicate
TxGemma
Designed to improve the efficiency of therapeutic development
ShieldGemma 2
Safety classifiers to detect harmful content in text inputs and outputs
CodeGemma
Intelligent code completion and generation across programming languages
T5Gemma
Encoder-decoder architecture to boost performance for deep contextual comprehension
PaliGemma 2
Trained at multiple resolutions to provide broad knowledge for transfer via fine-tuning
DataGemma
Integrates retrieval techniques to ground responses in real-world data
RecurrentGemma
Novel recurrent architecture for faster processing of long sequences
Gemma 3
The best open model that fits on a single NVIDIA H100 GPU
Build AI that helps people
The Gemmaverse
From cutting-edge techniques to personalized models, discover how developers are using Gemma open models.
Democratizing AI with offline access
Lentera runs an offline AI microserver for educators and students
Developing sovereign AI
Crane AI Labs improves Swahili language understanding in a lightweight model
Unlocking interspecies communication
Marine biologists and AI engineers partnered to develop DolphinGemma
Get started with Gemma
Platforms & integrations
Developer resources
Official documentation, quickstarts, and guides to build applications with Gemma
Get support
Ask questions on the developer forum or the Google Developers Discord #gemma channel
Large Language Models (LLMs), such as Gemma, may sometimes provide inaccurate or offensive content that doesn’t represent Google’s views.
Use discretion before relying on, publishing, or otherwise using content provided by LLMs.
Don’t rely on LLMs for medical, legal, financial, or other professional advice. Any content regarding those topics is provided for informational purposes only and is not a substitute for advice from a qualified professional.