OpenAI's Q2 revenue was approximately $6.7 billion, up 18% year-over-year, but operating losses widened to $12.3 billion, compared to about $9.3 billion last quarter. Anthropic's Q2 revenue was around $11.5 billion, with adjusted operating profit turning positive. The gap between the two might be due to the completely different paths the companies are taking. OpenAI feels more like it's building an AI platform: ChatGPT for the consumer market, enterprise services, agents, next-gen model training, and infrastructure are all being pushed forward simultaneously. The money it's burning now is more about securing the AI entry point for the next decade. Anthropic, on the other hand, seems more like a high-end enterprise AI company. Claude is rapidly penetrating scenarios like coding, finance, consulting, and enterprise automation—especially Claude Code, which integrates directly into developer workflows. Enterprise clients are willing to pay for tools that save them dozens of hours of work, rather than just buying a chatbot. So, in the next phase of AI competition, beyond model parameters, the key will be cost-effectiveness—how to get more computing power at the lowest cost. Model capabilities determine the ceiling, while the business model determines cash flow. OpenAI might be betting on the future in advance, while Anthropic is proving that AI can quickly become a profitable business.