[OpenAI Introduces New Optimization Plan, Reduces Model Inference Costs by Over 50%] BlockBeats News, June 30, according to a report by The Information, OpenAI engineers have developed a series of new model optimization techniques, reducing model inference costs by over 50% and decreasing reliance on NVIDIA GPUs. The report states that OpenAI may potentially use part of the cost savings in the future to lower API service prices or increase user query limits for products like ChatGPT.