If you want to use non-OpenAI models in the Next Web client, follow these steps.
First, ensure you have upgraded to the latest version (v2.13.0). In this version, a method for custom models has been added:
Enter Custom Model Name:
Since the Next Web client automatically determines the model provider based on the input model name, and the key provided by this site is only applicable to the OpenAI interface.
Therefore, when entering a custom model name, the format should be: “+‘desired model name’@OpenAI”.
For example, to use the gemini-pro model, enter the custom model name as: “+gemini-pro@OpenAI”.
This way, you can manually direct the model provider to OpenAI.
Confirm Selection:
In the model selection interface, confirm that the manually added model (OpenAI) appears below.