Tech
OpenAI Introduced the GPT-4o Mini Model
OpenAI has introduced the GPT-4o Mini model, the latest innovation in the field of artificial intelligence. This powerful and compact model stands out with its natural language processing capabilities. It offers features that will revolutionize applications in artificial intelligence.
OpenAI Introduces New GPT-4o Mini Model
OpenAI has introduced a smaller and more affordable version of its high-level large language model that powers ChatGPT. This new model, called GPT-4o mini, will offer developers a 60% lower cost for creating AI-powered applications and services compared to GPT-3.5 Turbo, which is the smallest model developed by OpenAI so far.
GPT-4o mini represents a significant advancement not only for developers but also for consumers. As of today, this model will replace GPT-3.5 Turbo for free ChatGPT users. This change means that your basic ChatGPT experience will be significantly improved.
OpenAI claims that the GPT-4o mini scored 82 points in an industry benchmark called MMLU (Measuring Massive Multitask Language Understanding). This benchmark consists of a total of 16,000 multiple-choice questions across 57 academic subjects. The score achieved is slightly lower than the 88.7 points scored by GPT-4o, but higher than the 70 points scored by GPT-3.5 Turbo.
AI experts warn that such benchmarks should not be relied upon to measure how intelligent AI systems are. However, as it stands, it continues to be one of the best methods we have for evaluating the performance of large language models. The smaller versions of high-level models give developers more flexibility in creating AI-powered applications. Considering that not everyone needs access to the largest and most powerful models for every use case, the advantages provided by smaller models are quite evident.
Similarly, Google announced its lightweight model called Gemini 1.5 Flash in May, which the company stated is optimized for speed and efficiency. Additionally, other AI companies such as Anthropic have started offering smaller versions of their full-scale models.
GPT-4o mini currently has the ability to process and generate text and images. However, it is believed that this model may also handle other types of content in the future, such as audio and video. While GPT-3.5 Turbo is distinct from ChatGPT, developers continue to access it through OpenAI’s API to create applications and services.