Google introduces the LLM model with the self-service Gemma 3, showcasing its prowess in the Chatbot Arena by outperforming DeepSeek-V3, Llama 3 405B, and o3-mini with its impressive 27B stature.
A key feature of Gemma 3 lies in its expansion of the context window to 128,000 tokens, allowing for a more extensive data processing capacity, including image inputs, bringing it closer to Gemini. Google touts the high capabilities of this 27B model, which can run on a single card and comes with a officially reduced resolution version.
It is now operational on Google AI Studio, ollama, and can even be utilized in a serverless manner via Cloud Run.
TLDR: Google has unveiled the Gemma 3 model with enhanced capabilities in the Chatbot Arena, showcasing its superiority over other models like DeepSeek-V3, Llama 3 405B, and o3-mini.
Leave a Comment