OpenAI moves forward with GPT-4.5 deprecation in API, triggering developer anguish and confusion

Words spread quickly through machine learning and the artificial intelligence community on the social network X yesterday: Openai was sending developers an email to notify them that the company will remove one of the largest and most powerful LLMS models, GPT-4.5, from the official Openai application interface (API) On July 14, 2025.

However, an Openai Venturebeat spokesman told email, the GPT-4.5 inspection will remain options for individual Chatgpt users through the form of the drop-down model in the upper part of the application.

But this means that any third-party developers have built applications or workflows above the GPT-4.5 inspection (we will call it GPT-4.5 from here on the sake of simplicity, given that the complete GPT-4.5 is not provided on this basic system), now you need to switch to one of Openai about 40 (!!!) is still available through API.

The news quickly spread to X, where developers and artificial intelligence lovers posted reactions ranging from disappointment to confusion.

Some described GPT-4.5 as a daily tool in their workflow, praising its accent and reliability. Others asked about the logical basis behind the launch of the model in the first place if it was short -term.

“This is a sad-GPT-4.5 is one of my favorite models,” Bumrahbachi Books.

Ben Hayak, co -founder of AI’s monitoring and performance monitoring platform, launched a “tragic” step, adding: “O3 + 4.5 are the models that I use every day.”

Another user asked FlowersLop frankly, “What is the purpose of this model all the time?”

The decline has been planned since April

Despite the strong reaction, Openai had already announced a plan for the GPT-4.5 inspection in April 2025 at the launch of the GPT-4.1.

At that time, the company stated that the developers will have three months to move away from 4.5. Openai frame the model as a experimental show that provides visions of future development, and said that he will continue to learn from GPT-4.5 to a repetition in the future-especially in areas such as creativity and writing of differences.

In the follow-up response to Venturebeat, Openai Communications confirmed that the email in June was just a tangible reminder and that there are currently no plans to remove the GPT-4.5 from Chatgpt, where the model remains available.

Society speculation of cost and model strategy

However, the neglect facing the developer leaves a gap for some users, especially those who have built a workflow or products on the specific GPT-4.5 properties.

Some of the society’s speculation that high account costs may affect this step, noting that similar changes occurred with previous models.

Others have pointed to the updates of the last application programming interface, including a significant reduction in the GPT-3.5 cost (referred to internally as O3), which is now 80 % less than before.

User @chatgpt21 commented that GPT-4.5 is “the best illogical model for opening artificial intelligence on all standards and is clear, and expected that Openai” will add test time, O3 will explode from water. In order to expand the TTC range to expand pre -training scope

The end of the road for GPT-4.5 via API- encouraged developers to immigrate to GPT-4.1

Openai directed the developers to its online forum to get questions about the deportation to the GPT-4.1 or other models. With the API closed to preview the GPT-4.5 defined in mid-July, the difference that depends on the model is now less than a month to complete this transition.