Assistants (OpenAI)

GPTs are custom models (like the assistants) but they are solely usable on OpenAI's website. Assistants let you create custom models and use them as you want, using an API like AI Engine does. You can of course use GPT-4 to create your own assistants. 😊

First, you will need to access the AI Engine dashboard and, under the "Server Modules" tab, enable the Assistants feature.

Notion image

This action will generate a new tab where you can view all of your OpenAI assistants.

Notion image

To create them, head to the OpenAI Assistants dashboard.

Notion image

Once you've created one or have some assistants ready, return to the AI Engine Assistants tab and click the refresh button in the top right corner. You should see it/them appear in the table.

Notion image

There's no need to return to this tab anymore. Now that the assistant is available, let's proceed to use it! 🐈

If you don't see your assistants after refreshing the list, please ensure that no errors are occurring. You can verify this by opening your developer's console and checking your PHP Logs. Additionally, make sure that your server can communicate with OpenAI and that your WordPress REST API is operational and you have the right permalinks configured.

For this, navigate to the chatbot tab where you operate your regular chatbots for AI Engine. Select or create a new chatbot, and in the Main Setting section, you will be able to change the chatbot mode to assistant.

Notion image
Notion image

And it's done! You are now using your OpenAI assistant directly within AI Engine. Have fun! 😊


Organizations’ assistants

You can connect to your OpenAI account and go to the API Keys section to select your default organization. If it's already the case, switch to personal, confirm the dialog box, and switch back to your organization to reset. Then, you'll need to create a new API Key so that the header matches. Once this is done, you can refresh the assistant in AI Engine, and you should see it appear! 😊

Did this answer your question?