Currently, the AI Settings page allows for model selection for each operation type. Text to image currently only has choices for dall-e-2 and dall-e-3. OpenAI just released much improved image generation support with GPT 4o model.
Go to the AI settings page here /admin/config/ai/settings , then navigate to text-to-image
Add ability to select GPT 4o as test-to-image model of choice.
Make GPT 4o the default choice in the text-to-image field if Open AI is enabled.
Active
1.1
AI Core module
Hi @andywaldrop,
The support of different functionalities for OpenAi models is provided by the OpenAI Provider → module.
The Open AI API doesn't support yet image generation with gpt-4o. You can see more details in the Open AI API supported models pageand the gpt-o4 model