Google recently introduced its latest AI model, Gemini, which is available in three versions: Nano, Pro, and Ultra. Gemini Ultra, the most powerful of the three, reportedly outperformed OpenAI's GPT in almost all benchmarks, according to insider sources.

Company insiders claim that Gemini Ultra outperforms ChatGPT in 32 of 30 widely used benchmarks for testing Large Language Models (LLM). This development positions Google to potentially close the gap with its AI competitor, OpenAI, known for its powerful GPT model that has gained great popularity in the industry.

Also read: Alphabet Reduces Costs for Its Advanced AI Model Gemini

As reported by Insider, Gemini achieved a remarkable 90% score in the Ultra Massive Multitask Language Understanding (MMLU) benchmark, making it the first model to outperform human experts. The MMLU assesses knowledge and problem-solving skills in 57 subjects, including mathematics, physics, history, law, medicine and ethics.

While the Gemini Nano and Pro models are already in use in Google products, the Gemini Ultra is scheduled for release in early 2024. Gemini Nano Pixel 8 Pro powers the AI features in the smartphone, Gemini Pro is integrated into the Bard chatbot, and it will also. Access to enterprise customers and AI Studio developers through Google's Vertex AI program.

Eli Collins, vice president of product at DeepMind, the Google division responsible for developing the AI platform, highlighted Gemini Ultra's ability to understand complex information in various formats such as text, images, audio, and code.

Collins mentioned that some of the data used to train the application was obtained from publicly available web content, although the specific origin was not clearly disclosed by the company.