How to Handle Text + Image Generation Together in same model Using AI SDK?

[viraj-7span]

Hi team,

I'm currently using generateText with gpt-4o via the @ai-sdk/openai package to stream or generate chat responses.

Now I want to also generate images (via DALL·E) based on user prompts like “generate image of a black dog”.

My questions:

1. 🔹 Can I generate images using the same generateText or gpt-4o model, or do I need to call something separately?
2. 🔹 What’s the best practice to handle both text and image generation together in one unified flow?

I saw there's a generateImage() function — should I trigger it manually when the user input looks like an image prompt? Or is there a cleaner way to do this using tool calling / function calling?

Any examples or patterns you recommend would be appreciated!

Thanks!

[Aarbel]

@viraj-7span did you manage to do it yourself ?