OpenAI启动Flex处理模式：API成本分配，但响应速度和稳定性已被牺牲

According to today's Foreign Media Techcrunch, to further increase the investment in competition along with generative competitors of AI such as Google, Openai announced the launch of a new API service of "FLEX processing mode" - users can receive slow speeding speed and "occasional unattended resources Openai said FLEX is currently opening testing on newly released O3 and O4-mini understanding models, mainly targeting "non-production" activities with lower priorities such as model review, data expansion,以异步处理等为例，flex模式费用为每百万个输入令牌5（在家中注：当前的汇率约为36.5 rmb，相当于Hurmb 750,000），而每百万个输出令牌为每百万美元的汇率（当前汇率）大约为40美元。每百万个输入令牌收取11美元将每百万个投入托图斯的投入量达到每百万美元0.55美元（现金利率约为每百万的产出代币161元。完成新推出的过程达到公司用户消费水平。