TMTPOST-- OpenAI, a leader in artificial intelligence, has announced that its most advanced model, GPT-4o, can now be fine-tuned and customized by developers for business applications. This new capability allows companies to enhance the model with proprietary information and adjust responses to meet specific needs, marking a new era in AI-powered solutions. In a blog post, OpenAI revealed that developers using the OpenAI API can fine-tune GPT-4o with their own datasets, enabling the model to perform tasks like suggesting products or responding to queries in a predetermined format, such as bullet points. OpenAI also noted that GPT-4o mini, a smaller and more cost-effective version of the flagship model, is also now available for fine-tuning. "From coding to creative writing, fine-tuning can have a large impact on model performance across a variety of domains," the company stated. One example of the potential benefits of fine-tuning is a partnership with Cosine, a Y Combinator-backed startup. By fine-tuning GPT-4o on real software engineering scenarios, Cosine’s AI-powered assistant, Genie, was able to mimic human engineers' problem-solving methods, outperforming other AI coding tools. OpenAI sees this as a significant step in improving business operations across industries through AI customization.