OpenAI went to an important structure change in GPT systems, which aims to offer personalized artificial intelligence experience. GPT versions developed by users with special settings and commands for specific tasks can now switch between different models. This way of use, which has been limited to the GPT – 4O so far, has started to support many models with the new update. This change offers significant flexibility, especially for users who need different performance levels in daily workflows.
Announced by OpenAI in late 2023 personalized gpt Its feature offered a structure where users could record repetitive commands. Thanks to this structure, users did not have to re -command the GPT versions for specific tasks and re -command every time. These GPTs, which are used in fields such as student studies, social media content preparation and document summarizing, have become a practical tool that saves time. However, their initially limited only one model, GPT – 4O, could not offer enough flexibility in some tasks.
With the new regulation, users now, GPT – 4O, GPT – 4.1, GPT – 4.5In addition, mini versions that require lower processing power of these models GPT – O4‑Mini And GPT -O4‑MİNİ – HIGH can choose between options such as. Each of these models offers different performance and process capacity. Especially in tasks that require low resource consumption, the preferred “light” models enables the efficient use of system resources. For tasks such as heavy analysis and long text production, more advanced versions can be used.
Model exchange is still limited in GPT versions that use action features
However, some limitations continue to be valid. If the personalized GPT version contains advanced features such as internet access, API calls or code running, this GPT’s model selection is still limited. In other words, only choice can be made between certain models. OpenAI states that he is aware of this restriction and has been working to remove it in the future. For now, GPTs with these features do not support wide model options.
In the model selection process, a recommended starter model for users can also be defined. This option, which is determined as the “Recommended Model ,, is automatically activated when each new session is started. If the user does not have the right to access this model, the system can automatically assign another model with similar features. In this way, the process can progress faster without having to make a manual choice in each session.
New model options can only be used by users with paid plans. Right now Pro, Plus and Team Access to this update can be provided through subscriptions. Enterprise And Edu For users, this feature has not yet been opened. OpenAI states that it plans to provide support for these user groups, but no clear explanation has been made about the timing.
The changes are a development that directly affects the general functioning of the system. Users can now determine the most suitable model according to the type of task. This shortens the processing time and provides a more economical use. Particularly the choice of mini models in task -oriented, short -term transactions offers a certain efficiency advantage.
The GPT personalization feature has been preferred by more advanced users. However, with the provision of model flexibility, it may be possible to try this feature by a wider audience. Creating a GPT defined according to a specific task set and testing with different models can increase usage scenarios.
Nevertheless, the expansion of the model selection feature may not be seen as a sufficient change alone. Particularly valid for GPTs that use features such as API integration and real -time data analysis may prevent some users from directing directly from this transition. However, it can be said that it is an important step taken in terms of general access ease.
Finally, this model variety offered by OpenAI offers a significant opportunity for users who want to create GPT versions that are lower costly and suitable for different usage needs. User feedback on which model is most suitable for which task can determine the future aspect of the system. These developments can prepare the ground for more efficient use of personalization feature.