Google Gemini is a powerful new family of large language models (LLMs) designed to be multimodal, meaning they can understand and process different types of information like text, code, audio, images, and video. It builds upon the previous LaMDA and PaLM 2 models, offering significant advancements in capability and safety.
Released in December 2023, Gemini comes in three versions: Ultra, Pro, and Nano. Ultra is the most powerful, tackling complex tasks, while Pro excels at scaling across various situations. Nano, the most efficient, runs directly on devices.
Key features include:
Google Gemini was previously known as Google Bard.
