
As a UI/UX designer, I particularly value Gemini's ability to work with both text and visual references within the same conversation. I use it to explore design directions, refine creative concepts, research visual ideas, and work through detailed image-editing instructions.
One workflow I found particularly interesting was refining a product image using specific design requirements. I could provide an exact HEX colour, describe the placement of branding elements, and specify which visual details, such as fabric texture, folds, and lighting, needed to remain unchanged. This makes the interaction feel more like communicating a design brief than simply entering a text prompt.
I also appreciate being able to iterate on an idea through follow-up instructions instead of rebuilding the entire request each time. For early-stage visual exploration and creative problem-solving, this gives me more room to experiment before moving into detailed design work. Review collected by and hosted on G2.com.
The main limitation I've noticed is that detailed visual instructions do not always translate into the exact result intended. When several constraints are involved, such as colour accuracy, object placement, typography, and preserving existing visual details, some elements may need additional refinement.
Longer creative conversations can also require occasional repetition of earlier instructions. I would like to see more consistent handling of reference images and better control over individual elements during image editing, particularly when precision matters. Review collected by and hosted on G2.com.