Google announced the launch of this highly capable and versatile AI model, designed to work in multiple dimensions (multimodal). This means it can understand and process various types of data, including text, code, audio, images, and videos.

Gemini 1.0 AI by Google: Gemini is Google's most powerful AI model, outperforming widely used academic benchmarks in 30 out of 32 criteria. Additionally, it is the most flexible model, capable of functioning across various devices, from data centers to mobile devices.

The initial version of Gemini 1.0 includes three models:

• Gemini Nano: A compact-sized model
• Gemini Pro: A medium-sized model
• Gemini Ultra: A large-sized model

Gemini is already being applied to some of Google's products, including bars and Pixel 8 Pro. Developers and organizational customers will be able to access Gemini Pro through Gemini's API in Google AI Studio or Google Cloud Vertex AI starting from December 13. Gemini Ultra is currently undergoing rigorous reliability and security checks and will be made available to a select group of testers and experts for trial use before being officially released to developers and organizational customers early next year.

Gemini has the potential to revolutionize various industries, from science and medicine to finance and education. It can be utilized to generate new deep learning data from large datasets, perform complex tasks automatically, and create innovative products and services.

Gemini is also being employed to develop advanced AI assistants, such as AlphaCode 2, capable of solving programming challenges that extend beyond conventional programming related to complex mathematics and theoretical computer science.

As of now, Gemini 1.0 supports only the English language, with support for other languages expected to be added gradually, according to Google's plans.

Gemini 1.0 marks a significant milestone in Google's AI development and may herald the beginning of a new era of powerful and capable AI.

Source: GizChina