Introducing Gemma 3, Google’s latest Compact AI Model designed for efficient AI application development. Available in various sizes, Gemma 3 is a powerful and portable open model that outperforms competitors on benchmarks.Caption:Google Gemma 3: Compact AI model illustration.
The Highlights:
Google has launched Gemma 3, the latest addition to their range of lightweight open models powered by Gemini 2.0, which is a Compact AI Model.
Gemma 3 comes in various sizes such as 1B, 4B, 12B and 27B and is considered Google’s most advanced and portable open model yet.
The Gemma 3 models are designed to run directly on phones, laptops, and devices to assist developers in creating AI applications that can analyze text, images, and short videos but respond only in text.
Google claims that Gemma 3 outperforms competitors like Meta’s Llama-4-5B model and OpenAI’s o3-mini and DeepSeek-V3 on various benchmarks while running on a single GPU. It also features a large context window of up to 128k tokens for processing vast amounts of information.
The tech company claimed these models are their “most advanced, portable and responsibly developed open models yet.”
Introducing Gemma 3: Google’s Cutting-Edge Compact AI Model
Google has recently unveiled Gemma 3, the latest addition to their lineup of lightweight open models powered by Gemini 2.0 models. In a recent blog post, Google announced that Gemma 3 is now available in various sizes such as 1B, 4B, 12B, and 27B.
The tech giant described these models as their “most advanced, portable and responsibly developed open models yet.”
Just like its predecessors, Gemma 3 has been designed to seamlessly run on phones, laptops, and other devices to assist developers in creating AI applications. These models support over 35 languages and can analyze text, images, and short videos while responding only in text format.
Google claims that Gemma 3 surpasses Meta’s Llama-4-5B model as well as OpenAI’s o3-mini and DeepSeek-V3 on various benchmarks while operating on a single GPU. With a 128k-token context window, Gemma 3 can efficiently process large volumes of information.
In addition to the release of Gemma 3, Google has also introduced ShieldGemma2 image safety classifier aimed at filtering out sexually explicit or violent images. These models are readily available for download through platforms like Kaggle, Hugging Face or Google Studio.
The launch of Gemma marks another milestone in Google’s commitment towards advancing AI technology for developers worldwide.
Published – March 13th, 2025 at11:02 am IST.
Gemma 3 is the latest addition to Google’s range of lightweight open models, powered by their Gemini 2.0 models. These compact AI models are designed for portability and can run directly on various devices like phones, laptops, and more.
Google claims that Gemma 3 surpasses competitors like Meta’s Llama-4-5B model and OpenAI’s o3-mini and DeepSeek-V3 on various benchmarks while running on a single GPU. With a 128k-token context window, these models can efficiently process large amounts of information.
In addition to Gemma 3, Google has also introduced ShieldGemma 2 image safety classifier to filter out explicit or violent images. These models are available for download through platforms like Kaggle, Hugging Face, or Google Studio.
Tanishq Raghvan, editor at TechVatsalya.com, specializes in AI, systems architecture, and emerging technologies, delivering expert insights and analysis.
