Mint turns photos into explorable 3D cities

Hot take: this is less about “AI image generation” and more about turning reference images into a navigable spatial experience, which is a much stronger demo if Mint can make the geometry feel believable.

• –The product’s core pitch is clear: generate 3D worlds or models from text/images, then edit and share them.
• –The demo leans on nostalgia and place familiarity, which is smart because it makes the output immediately legible to a broad audience.
• –The post is light on technical proof, so the real question is consistency: can Mint preserve recognizable city structure across more than just a few showcase scenes?
• –As a growth angle, “visit any city at any point in time” is a much better story than generic 3D generation.