OpenAI has significantly reduced the inference costs of its AI models, including ChatGPT, by more than half through targeted optimizations. According to The Information, these improvements have also allowed the company to decrease the number of Nvidia GPUs required to operate ChatGPT to just a few hundred at times.
This cost reduction reflects OpenAI’s ongoing efforts to improve the efficiency and scalability of its AI services, which could have broad implications for the deployment of large language models across industries.
For the Japanese market, where AI adoption is rapidly growing, such advancements may lower barriers for businesses integrating AI-powered solutions into their operations, potentially influencing equity and tech sector dynamics.