Use GPT-4o today without updating your app.

AIProxy customers, you can use GPT-4o in distributed versions of your apps right now, without submitting an app update. The Model Overrides feature, found in the developer dashboard, allows you to dynamically change the model string contained in API requests that route through the proxy. This means that requests from your app using GPT-4 Turbo can be upgraded to GPT-4o without taking out an app release.

The Model Overrides features is particularly helpful on days like today, when an AI provider releases a new model. In this case, OpenAI released GPT-4o, which has speed and pricing improvements over GPT-4 Turbo, as well as multi-modal improvements.

Turn on model overrides by navigating to the 'Model Overrides' tab in the AIProxy sidebar:

Once you apply an override, you will see the model swapped out in the live console when you make a request: