Last night, Openai released its latest artificial ielligence model, O3-MINI. This model can be used. The poi is that Openai for the first time provides one of its “reasoning” models to ChatGpt users for free. To experience this model, just select the Reason button in the message writing section and start the job.
According to OpenAI experts, the O3-Mini model has a higher speed and accuracy than its predecessor, O1-Mini. The results of the A/B tests performed by the company show that O3-mini operates 24 % faster than O1. Also, this new model can be closer to the performance level of the O1 by adjusting to the “moderate” reasoning mode, in some performance measureme tests in mathematical, programming and science. Similar to other OpenAI rational models, the O3-Mini model shows how the user gets the answer, not just to answer the request. Another promine feature of this model is its compatibility with the ChatGPT search, which allows the web search to find the latest information and useful links. Openai has announced that it is trying to iegrate the search feature in all its reasoning models.
“The launch of the Openai O3-Mini model is another step towards the Openai mission to expand the boundaries of affordable ielligence,” Openai said in a stateme. “We facilitate access to quality artificial ielligence by optimizing reasoning in the STEM areas and at the same time reducing costs.” “This model coinues our successful trend in reducing artificial ielligence costs-so that since the iroduction of GPT-4, we have reduced pricing per token by 95 %-while maiaining high level reasoning abilities. We’ve done. “With the expansion of artificial ielligence applications, we are still committed to leading in this emerging field and developing models that balance ielligence, efficiency and large -scale safety.”
With today’s announceme, the O3-MINI model will replace O1-MINI in the model selection section. In addition, OpenAI has increased the use limit for Plus users and the ChatGPT team from 50 daily messages with O1-Mini to 150 daily messages with O3-Mini. Also, the Pro Plan, recely iroduced by Openai for $ 200 a moh, offers unlimited access to the new system.
When OpenAI first iroduced O3 and O3-MINI models late last year, CEO Sam Altman said the second model would be released “almost in late January.” On January 17, Altman released a more accurate schedule by posting on the X -Social Network, announcing that Openai “plans to launch it within the next few weeks.”
Now that O3-Mini is available, it can be said that its release has been accompanied by an urge seasoning. On January 20, with Altman at the inauguration of Donald Trump, the Chinese company Deepseek was quietly released. Uil January 27, the company’s chats surpassed ChatGPT as the most downloaded free app in the United States AppStore. The sudden success of Deepseek led to a $ 1 trillion decrease in the stock market value, and it definitely put Openai in a surprise position.
26,100,000
24,199,000
Toman
Source: Engadget
