OpenAI Announces Its New AI Model

OpenAI recently announced a new model of its software, “GPT-4o,” that will provide users with unique capabilities.

OpenAI experts noted that the new AI model, "GPT-4o mini," is a simplified version of the GPT-4o model, outperforms many previously tested AI models, allows for many tasks to be performed at a low cost, and is suitable for use with many applications to provide fast performance.

GPT-4o mini supports text and graphical interfaces, will soon provide support for video and audio content, and can process up to 128,000 input symbols and up to 16,000 output symbols per request.

Thanks to the advanced software techniques used by OpenAI, the new model can also process texts written in languages ​​other than English very well.

According to the available information, GPT-4o mini outperforms many competing AI software. For example, this model achieved 82% in reasoning tests, while Gemini Flash achieved 77.9%, and Claude Haiku achieved 73.8%. As for tasks related to mathematics and programming, the new “mini” model scored 87% in MGSM, compared to 75.5% for Gemini Flash and 71.7% for Claude Haiku.