Google has recently introduced Gemini, a collection of AI models designed to cater to various needs from lightweight mobile applications to heavyweight enterprise solutions. Gemini comes in three versions: Gemini Nano, Gemini Pro, and Gemini Ultra, each serving specific purposes and use cases.

Gemini Nano is the lightweight version of the Gemini AI model, specifically tailored for native and offline use on Android devices. It empowers mobile applications with on-device AI capabilities, ensuring faster response times and increased privacy. Pixel 8 Pro users will benefit from new features powered by Gemini Nano, highlighting its practical application in enhancing user experiences on mobile devices.

Gemini Pro, a more robust version of the model, is set to power numerous Google AI services. It has already been integrated into Bard, Google's language model, to enhance advanced reasoning, planning, understanding, and language comprehension. Starting December 13th, developers and enterprise customers can access Gemini Pro through Google Generative AI Studio or Vertex AI in Google Cloud.

Early next year, the tech giant plans on introducing 'Bard Advanced', which will leverage the even more powerful Gemini Ultra. As per Google executives, Gemini Ultra will be the most significant update to Bard -  standing out as the most capable model in the Gemini lineup, designed for data centres and enterprise applications. Its capabilities are expected to significantly impact large-scale AI implementations, offering "unprecedented performance and efficiency for complex tasks".

As per Sundar Pichai, CEO of Google, Gemini is being positioned as the future of Google, with plans for integration into various facets of the company's ecosystem, such as the Google search engine, ad products, the Chrome browser, etc. While the AI model suite is currently only available in English, Google plans on launching support for other languages soon.

According to Google DeepMind CEO Demis Hassabis, Gemini has beaten ChatGPT in 30 out of 32 benchmarks for testing AI-based capabilities. According to him, while most of the benchmarks were closely won by Gemini, the advantage of Gemini was being able to better decipher audio and video inputs.