Open AI launches new GPT-4o Mini model
The company called the new release “the most capable and cost-efficient small model available today”.
OpenAI is releasing the new, more accessible and simpler version of its GPT-4o AI model in order to attract more developers and corporate clients.
According to the company's report, GPT-4 has improved its audio, video, and text generation abilities, and now it can process up to 50 different languages, with improved speed and accuracy. Over time, the new model will be able to process other types of content.
The GPT-4 model will completely replace GPT-3.5 Turbo, but the older model will still be available for developers through the API for a period of time.
OpenAI has also announced that GPT-4o mini is the company's first AI model that uses a new security technique developed by them called "instruction hierarchy". This approach aims to make the AI system prioritize certain companies instructions over others, making it more difficult for users to make the tool perform tasks it shouldn't.
The company has announced that the updated model is now available for free users and subscribers of ChatGPT Plus and Team. It will also be offered to corporate customers starting next week.
The GPT-4 model will completely replace GPT-3.5 Turbo, but the older model will still be available for developers through the API for a period of time.
