Loading…
Y2K is a list of specific objects, not a filter. A bandana, narrow sunglasses with tinted lenses, big hoop earrings, layered chains and a halter top: remove one item and the look collapses back into an ordinary street selfie. The template assembles the full set.
In frame: a bandana in a black-and-white paisley print, a small braid running along the face, sunglasses with orange lenses, gold hoop earrings, a black halter top with corset gathering at the hem, and several chains of different lengths. The shot is a selfie taken from a high angle in bright natural light, with asphalt as the background. One line of the prompt forbids altering the face; another asks for realism rather than stylisation.
Anyone who wants a recognisable 2000s look without tipping into cosplay: the tinted lenses and the bandana supply the era, while the asphalt and the high angle keep the frame in ordinary-selfie territory. It reads equally well as an avatar and as a post.
Since October 2025 it has been run 181 times by 139 different people, and all 181 runs finished with a delivered image. Five results were published by their authors to the public feed.
One picture. The bandana, the sunglasses, the jewellery and the top come from the template. It helps if:
The Retro Glam category holds ten styles, and this is the first of them with a description. Its closest neighbour by era from another category is the vintage polaroid: there the era comes from the medium — the film border and the grain — while here it comes from objects and accessories, and the picture stays digital and sharp. It differs from the fashion portrait in being a street selfie shot at arm's length rather than a studio frame.
Model nano-banana-2, vertical frame, median 34 seconds (measured across 42 runs with a recorded completion time), 200 credits. Custom prompts are not accepted: the set of objects is fixed, because the set of objects is the style.
The visual code of roughly 1999–2005: bandanas and headscarves, narrow sunglasses with tinted lenses, large hoop earrings, layered chains, halter tops with corset detailing. This template assembles that exact set.
No. Every object is written into the prompt and appears on its own. All that is needed is one portrait, ideally shot in daylight so the lighting matches the scene.
The prompt contains an explicit instruction not to change the face. That said, the lenses cover the eyes, so the likeness rests on the lower half of the face — a larger, sharper source gives a better result.
The median is 34 seconds across 42 runs with a recorded completion time. All 181 runs of this style finished with a delivered image.
