API option that offers a lower artificial intelligence model usage prices than OpenAI: Flex processing

OpenAI, we can translate as flexible processing Flex processingreleases. Flex processing, which is currently in the beta version for O3 and O4-Mini, is an API option that provides lower artificial intelligence model usage prices in exchange for the lack of resources from time to time. In this context, Flex processing offers more slow response times.

Flex processing is planned to be used in lower priority and non -production tasks. Among these tasks; Model evaluations, data enrichment and asynchronous workflows. According to shared information; Flex processing API reduces costs by half. Flex processing for O3, about 750 thousand words from a million input token to Token 5 dollars per Million and Million Output token to Token 20 dollars per. At this point, the standard processing is 10 dollars per million input token and a million output token is 40 dollars per per hand. Flex, the price of the O4-Mini to the million input token 1.10 dollars per and million output token to 4.40 dollars per million input marker-per-selection of 0.55 dollars per to the token to 2.20 dollars per Token.

OpenAI said that developers at 1 to 3rd levels of the usage level hierarchy will have to complete the newly introduced authentication process to access O3. The stages are determined according to the amount of money spent for the OpenAI services. If you remember, we have shared with you that OpenAI will need to authenticate to access future artificial intelligence models. Developers have to authenticate the reasoning summaries of O3 and other models to access the support of the Flow API support. O4-Mini can be used by customers at the expenditure levels from 1 to 5.

OpenAI offers a solution to its competitors that offer more budget -oriented models with Flex processing. On Thursday, Google’s Lower Input Token’s cost with Deepseek R1 in terms of performance with the cost of the reasoning model Gemini 2.5 Flash launched the Flash, OpenAI may have mobilized.