OpenAI's financial health raises important questions about the profitability of AI models, particularly GPT-5. Despite generating $6.1 billion in revenue over four months, the cost of running the model and its associated services exceeded that figure. OpenAI's inference compute costs alone were estimated at $3.2 billion, while total operating costs, including salaries, marketing, and administrative expenses, reached approximately $6.8 billion. This resulted in an operating loss of around $0.7 billion, even before accounting for R&D investments. The analysis highlights a distinction between gross profit and overall profitability. With a gross margin of about 48%, OpenAI appears to generate enough revenue to cover direct operational costs. However, when other expenses are included, the picture shifts to one of financial loss. The situation is compounded by a revenue-sharing agreement with Microsoft that takes about 20% of OpenAI's earnings. Although this partnership provides some financial relief through shared revenue, it still poses challenges for OpenAI's path to profitability. Estimating R&D costs for GPT-5 is complex, with OpenAI's total R&D expenses pegged at $16 billion for 2025. The article suggests that the model's short lifespan and the rapid pace of competition could hinder its ability to recoup development costs. While gross margins provide a more optimistic view of potential profitability, the reality is that sustained revenue generation may not be guaranteed, especially given the fast-evolving AI landscape.