Google launches Gemma 3 AI models that run on single GPU as company takes on Facebook’s Llama, DeepSeek and OpenAI

TOI Tech Desk / TIMESOFINDIA.COM / Mar 12, 2025, 21:00 IST

Text Size

Small
Medium
Large

Google introduced Gemma 3, a new series of open-source AI models designed for superior performance and accessibility, in a bid to compete with AI giants like OpenAI and Facebook. Gemma 3 supports over 35 languages with advanced text and visual capabilities, enabling developers to create powerful AI applications across various devices.

Google launches Gemma 3 AI models that run on single GPU as company takes on Facebook’s Llama, DeepSeek and OpenAI

has launched Gemma 3, a family of open-source artificial intelligence (AI) models designed for high performance and accessibility. The announcement comes as the company boosts its effort to better compete with companies including ChatGPT-maker OpenAI, Facebook and DeepSeek.
According to the company, Gemma 3 models build upon the success of the previous AI models, which recently celebrated its first anniversary with over 100 million downloads. Gemma 3 aims to empower developers to create powerful AI applications that can run directly on devices, from smartphones to workstations.
“Gemma 3 delivers state-of-the-art performance for its size, outperforming Llama-405B, DeepSeek-V3 and o3-mini in preliminary human preference evaluations on LMArena’s leaderboard. This helps you to create engaging user experiences that can fit on a single GPU or TPU host,�?the company said.

Gemma 3 is powered by same tech as Gemini 2.0

Gemma 3, derived from the same research and technology powering Google's Gemini 2.0 models, is available in various sizes (1B, 4B, 12B, and 27B) to cater to diverse hardware and performance requirements.
As per Google, the models outperform competitors in preliminary human preference evaluations, and offer advanced text and visual reasoning capabilities, enabling developers to build applications that analyse images, text and short videos.
Key features of Gemma 3 include out-of-the-box support for over 35 languages and pretrained support for over 140 languages, a 128k-token context window for processing and understanding extensive information and upport for function calling and structured output to automate tasks and create agentic experiences.
Gemma 3 integrates with popular development tools and platforms, including Hugging Face Transformers, Ollama, JAX, Keras, PyTorch, Google AI Edge, UnSloth, vLLM, and Gemma.cpp. Developers can easily customise and deploy Gemma 3 using platforms like Google Colab, Vertex AI, and NVIDIA GPUs, which have been optimised for maximum performance.

About the Author

TOI Tech Desk

The TOI Tech Desk is a dedicated team of journalists committed to delivering the latest and most relevant news from the world of technology to readers of The Times of India. TOI Tech Desk’s news coverage spans a wide spectrum across gadget launches, gadget reviews, trends, in-depth analysis, exclusive reports and breaking stories that impact technology and the digital universe. Be it how-tos or the latest happenings in AI, cybersecurity, personal gadgets, platforms like WhatsApp, Instagram, Facebook and more; TOI Tech Desk brings the news with accuracy and authenticity.

End of Article