The Wordian distribution of machine learning and the AI community within the social network X yesterday: Openai sent developers an email by which she notified that the corporate can be faraway from the official Openai application programming interface (API), API) on July 14, 2025.
As an Openai spokesman Venturebeat announced by email, the preview of the GPT-4.5 preview for individual chatt users via the “Dropdown model selection” menu stays an option at the highest.
But it means Openais almost 40 (!!!) different model offers which might be still available via the API.
The news quickly spread to X, where developers and KI enthusiasts publish reactions that range from disappointment to confusion.
Some described GPT-4.5 as day by day tools of their workflow and praised his tone and reliability. Others primarily questioned the justification behind the beginning of the model if it could only be short -lived.
“This is sad-GPT-4.5 is one in all my FAV models,” wrote @Bumrahbachi.
I’m HyakThe co-founder of the AI observability and performance monitoring platform Raindrop.ai, which is described and adds: “O3 + 4.5 are the models that I take advantage of the day by day.”
Another user, @Flowersslop asked bluntly: “What was the aim of this model on a regular basis?”
The shutdown had been planned since April
Despite the strong response, Openai had already announced the plan in April 2025 through the GPT-44.1 start to scale back the GPT-4.5 preview.
At this point, the corporate stated that developers had three months to alter from 4.5. Openai framed the model as an experimental offer that provided insights into the long run development and said that GPT-4.5 can be insight into future iterations-especially in areas resembling creativity and writing of nuances.
In a follow-up response to Venturebeat, Openai Communications confirmed that the e-mail in June was only a planned memory and that there are currently no plans for removing GPT-4.5 from chatt subscriptions by which the model stays.
Community speculation about cost and model strategy
Nevertheless, the developer devaluation leaves some users, especially those that had built up workflows or products around the precise features of GPT-44.
Some speculated in the neighborhood that prime calculation test costs can have had the move, and located that similar changes had occurred in previous models.
Others referred to the present API price updates, including a bigger cost reduction for GPT-3.5 (internally known as O3), which is now 80% lower than before.
user @Chatgpt21 Commented that GPT-4.5 is the “best non-argumentation model for open AI for all benchmarks and obvious”, and predicted that Openaai “add the test time to blow O3 from the water. To scale the TTC, it’s good to scale the pre-training”.
Encouraged the top of the road for GPT-4.5 via API developers to migrate to GPT-4.1
Openai has instructed developers in his online forum to migrate questions on migration about GPT-4.1 or other models. After the API was discontinued for the GPT-4.5 preview for mid-July, teams that depend on the model now have lower than a month to finish this transition.
