Remove anything from a photo

Object removal used to require careful manual masking in Photoshop's content-aware fill, and even then the result often had obvious patches or repeated textures. Modern image-edit models like Nano Banana handle the whole pipeline: they understand the scene semantically, identify what you want removed from your text description, and synthesise plausible content for the gap that matches the surrounding lighting and perspective. Common use cases include removing strangers from vacation photos, removing distracting background elements (lamp posts, parked cars, signs) from architecture and product shots, removing watermarks from your own previously-watermarked images, removing logos or text overlays, and removing ex-partners from photos you want to keep but not their face. The text-prompted approach is far easier than masking: instead of carefully tracing around the unwanted element, you just say 'remove the person on the left in the red jacket'.