FREE TOOL · BETA
Where was this photo taken?
Upload a photo. The tool reads its metadata, studies the visible clues, and ranks the places it could be — with the evidence for and against each one, and an honest “couldn't determine” when the evidence is not there.
Your photo stays yours
- Processed in memory and discarded when the analysis finishes — never stored, on any disk, anywhere.
- Never sent to a reverse-image search or any photo-matching third party.
- Place lookups send text clues only (a landmark or place name) to OpenStreetMap — never your image.
- No face recognition, ever. Results for homes are shown as a general area only, and no result is displayed tighter than 250 m.
Drop in a photo and this tool works the way a careful human would. First it reads the file itself: many camera and phone photos carry GPS coordinates in their EXIF metadata, and when they do, that is the strongest evidence there is. Then it studies the scene for clues — readable signs and place names, named landmarks, licence-plate styles, the side of the road traffic drives on, vegetation and architecture — and matches the named places against OpenStreetMap.
What you get back is not a confident-sounding guess. Every candidate is ranked by the evidence behind it, shown with the reasons for and against, a plain-language confidence level and an uncertainty radius. Sometimes the honest answer is “we couldn't determine this one” — when that happens the tool says so, shows the clues it did establish, and tells you what would help. It never invents a location.
This is a beta, and it has real limits. It does not run reverse image searches, it does not compare your photo against street-level imagery, and photos whose metadata was stripped by a messaging or social app — most copies you have been sent — are much harder to place than originals from the camera.
How it works
- 1Choose or drop a photo (JPEG, PNG or WebP, up to 10 MB). It is processed in memory and discarded when the analysis finishes — never stored.
- 2The tool reads the photo's EXIF metadata, including any embedded GPS coordinates.
- 3A visual pass extracts clues only — text, landmarks, plates, road layout, scene type. It does not decide the location.
- 4Named landmarks and place names are matched against OpenStreetMap, and every match is scored on its evidence.
- 5You get ranked candidates with confidence, an uncertainty radius and the evidence for and against — or an honest “couldn't determine”.
- 6Point results link onward: live cameras near the spot, and an “Explore location in 3D” jump onto the WorldEye globe.