OpenAI launches in a bid for more aggressive competition with rival AI companies such as Google. OpenAI has released the o4-mini reasoning model in beta. It is designed for lower-priority tasks, such as data enrichment, asynchronous workloads and model evaluations. It cuts API costs in half. Flex processing costs $5 per million tokens (750,000 word) and $20 per millions output tokens for o3, compared to the standard $10 input tokens and $ 40 per million tokens. Flex processing brings down the price for o4 mini to $0.55 per 1,000,000 input tokens. This compares to $1.10 per 1,000,000 input tokens.
The launch comes as the cost of frontier AI continues its upward trend and rivals release more affordable, budget-friendly models. Google launched its new Google Assistant on Thursday. Gemini 2.5 Flash () is a reasoning model which matches or beats DeepSeek R1 in terms performance and input token cost.
The ‘In an OpenAI announced the launch of Flex Pricing in an emailto customers. It also stated that developers from tiers 1 to 3 of its usage hierarchy would have to complete a newly introduced ID Verification process to gain access to o3. The amount of money you spend on OpenAI services determines the tiers. Verification is also required to access O3’s (and other models’) reasoning summaries, and the streaming API. OpenAI said earlier that ID verification was intended to prevent bad actors from violating their usage policies.
Kyle Wiggers, TechCrunch AI Editor. His writings have appeared in VentureBeat, Digital Trends and a variety of gadget blogs, including Android Police and Android Authority, Droid-Life and XDA-Developers. He lives in Manhattan, with his music therapist partner.
View Bio
