Join the event entrusted by business leaders for almost two decades. VB Transf transform gathers the people who build a real strategy of AI. Get more information
Word spread rapidly through the learning of machines and the AI community on the social network X yesterday: Openai sent to developers an email that would notify them that the company would eliminate one of its larger and powerful large language models (LLMS), GPT-4.5 Preview, of the official Interface of Openai Application (API). On July 14, 2025.
However, as Ventubbeat said in VentubeBeat by email, the preview GPT-4.5 will remain an option for individual chat users through the drop-down models menu at the top of the application.
But Stil means that all third-party developers who created applications or working flows in view GPT-4.5 (we will call GPT-4.5 From here to Out by simplicity, since a complete GPT-4.5 did not make available to this platform), now it is necessary The different offers of models of about 40 (!!!) of Openai still available in the API.
The news rapidly spread to X, where developers and AI fans published reactions ranging from disappointment to confusion.
Some described GPP-4.5 as a daily tool in their workflow, praising their tone and reliability. Others questioned the reason for launching the model first if it would be short -lived.
“This is sad: GPT-4.5 is one of my favorite models,” he wrote @Bumrahbachi.
I’m HyakThe co -founder of the Raindrop.AI platform of the AI and Performance Control Platform, called “tragic”, adding: “O3 + 4.5 are the models that I use most every day.”
Another user, @flowersslop Asked clearly, “What was the purpose of this model?”
Deprection had been planned since April
Despite the strong reaction, Openai had already announced the plan to waste the preview GPT-4.5 in April 2025 during the launch of GPT-4.1.
The company then stated that the developers would have three months to go from 4.5. Openai framed the model as an experimental offering that provided information for future development and said that it would bring GPT-4.5 learning to future iterations, especially in fields such as creativity and writing nuance.
In a follow-up response to VentubeBeat, Openai Communications confirmed that the June email was simply a scheduled reminder and that there is currently no plan to eliminate GPT-4.5 of Chatgpt’s subscriptions, where the model remains available.
Community speculation on the Cost Strategy and Model
However, developers’ oriented depretion leaves a gap for some users, especially those who had created workflows or products around the specific features of GPT-4.5.
Some of the community speculated that high costs of calculus may have influenced the movement, saying that similar changes had occurred with previous models.
Others referred to the recent API prices updates, including a significant reduction in the cost by GPT-3.5 (called internally as O3), which now has a price of 80% lower than before.
User @Chatgpt21 He commented that GPT-4.5 is the “best model of non-reasoning for the open AI at all points of reference and is obvious”, and predicted that Openai “will add the test time to be calculated O3 out of the water. To climb TTC that you need to heat the previous training”
The end of the road by GPT-4.5 through API-The developers encouraged to migrate to GPT-4.1
Openai has directed developers to their online forum for migration questions to GPT-4.1 or other models. With the stoppage of the API for pre-view GPT-4.5 established by mid-July, teams that trust the model now have less than a month to complete this transition.
Source link
