**[Video | OpenAI Board Chair Discusses Kimi K3: Open-Source Models Aren't Necessarily Cheaper]** On July 20, OpenAI Board Chair Bret Taylor, in an interview with CNBC, was asked about his views on China's newly launched open-source model, Kimi K3. He stated that when companies choose AI models, their primary concern isn't whether the model is "open-source" or "closed-source," but rather whether the token costs invested can deliver sufficient practical value. In his opinion, open-source models aren't necessarily cheaper, as they may require consuming more tokens to complete the same task. In contrast, cutting-edge models like those from OpenAI and Anthropic, while having a higher per-token price, are also more efficient and capable. Therefore, determining whether a model is cost-effective shouldn't focus solely on the token unit price but also on the total cost of completing tasks and the final outcomes.