GeoFinderAI

GeoFinderAI estimates the location of a photo by reading visual clues such as signs, buildings, plants, terrain and sky. It returns a region, country, city and coordinates with a confidence score, which is useful for curious users and open-source investigators.

The analysis runs in stages, starting with a vision model that lists location clues before scoring candidate places, and a typical answer takes under 30 seconds. The company says the model was trained on geotagged images from more than 190 countries and that its second version is more accurate than the first.

There is no subscription. New users get three free guesses, then buy token packs where each guess costs 10 tokens, and the larger packs add saved history and priority processing.