OpenAI has made a significant move by releasing a more affordable version of its popular AI model, GPT-4o. The lite model, called GPT-4o mini, is aimed at making OpenAI’s AI technology more accessible and cost-effective for users. It comes with a price tag that is 60% cheaper than its predecessor, GPT-3.5 Turbo.
GPT-4o mini has been designed to be more climate-friendly and will be made available to free users of ChatGPT, paying Plus and Team subscribers from July 18. Enterprise users can expect access to the model in the following week.
OpenAI claims that GPT-4o mini outperforms its competitors on major benchmarks. The model achieved an impressive score of 82% on the Massive Multitask Language Understanding (MMLU) test, which measures the intelligence and reasoning capabilities of language models. This score indicates that GPT-4o mini has a high proficiency in understanding and using language in real-life contexts.
Additionally, the lite model comes with new safety features to resist jailbreaks and prompt injections, enhancing the reliability and safety of its responses in large-scale applications. Initially available as a text and vision model in the API, OpenAI plans to expand its support to include text, image, video, and audio inputs and outputs in the future.
With GPT-4o mini, OpenAI has taken a significant step towards making their AI technology affordable and accessible to a wider range of users, while still delivering impressive performance and safety measures.