whereisthis.place uses two location paths. First, it reads EXIF and XMP metadata in the browser, showing embedded GPS coordinates immediately without uploading the original image or consuming AI credits. If coordinates are absent, visual analysis examines architecture, signage, road markings, vehicles, vegetation, terrain, and infrastructure. The system returns up to five ranked predictions with descriptions, confidence scores, estimated coordinates, and reasoning. Nominatim resolves predictions into place names and administrative boundaries, while an interactive Mapbox satellite map displays the results. Supported uploads include JPG, JPEG, PNG, WebP, and iPhone HEIC. Original photos are processed in memory and discarded after analysis; logged-in users may retain only a thumbnail and result JSON in search history. Results should be independently verified, especially for generic or low-detail scenes.