That loss includes the costs to train the future models.
Like Dario/Anthropic said, every model is highly profitable on it's own, but the company keeps losing money because they always train the next model (which will be highly profitable on it's own).
But even if you remove R&D costs, they’re still billions of dollars short of profitability. That’s not a small hurdle to overcome. And OpenAI has to continue to develop new models to remain relevant.
OpenAI "spent" more on sales/marketing and equity compensation than that:
"Other significant costs included $2 billion spent on sales and marketing, nearly doubling what OpenAI spent on sales and marketing in all of 2024. Though not a cash expense, OpenAI also spent nearly $2.5 billion on stock-based equity compensation in the first six months of 2025"