OpenAI has announced that it will reintroduce its $200 Pro subscription service tomorrow, while adjusting the API usage quota at this price point to half of the original plan. The new subscription service removes the previous 5-hour usage limit, allowing users to freely allocate their weekly quota. This adjustment is made possible by significant improvements in the inference efficiency of the underlying models and a tiered reduction in API pricing. Additionally, the new models GPT-6Sol and GPT-6Luna will be launched this week at a 50% discount off their original prices. OpenAI is translating technological advancements into cost-effectiveness for end-users, narrowing the cost gap between fixed subscriptions and pay-as-you-go options. The new service will also introduce several new privilege features that do not consume data quota. This strategic adjustment marks a substantial shift in OpenAI's billing logic for large model commercialization, reflecting that leading AI companies are driving industry evaluation standards from mere computational power usage to actual work efficiency. In the future, the core competition in large model subscription services will focus more on exclusive feature ecosystems and comprehensive application experiences.