Exploration · Visual AI

Locus

Where was this photo taken?

Locus upload screen: Where was this taken? Drop a photo here.

Drop in a photo and Locus gives a best-guess location on an interactive map. It uses the photo's own GPS data when it can, and when it can't, it says so with an honest uncertainty circle instead of a falsely precise pin.

Precise when it can be. Honest when it can't.

Fixed radius when the photo has GPS
1.7km
Radius for a visual guess
10–30km
Vision providers supported
3
API keys needed for GPS photos
0

How it works.

Metadata first, visual clues second, and a map that shows how sure the answer is.

Step 1

Read the file.

The browser parses EXIF data with exifr and converts HEIC photos when it can. Photos with GPS skip the AI entirely.

Step 2

Find the place.

GPS coordinates are reverse-geocoded with OpenStreetMap. Without GPS, a compressed image goes to a vision model that reads architecture, vegetation, signs, vehicles and terrain.

Step 3

Show the uncertainty.

The result appears on a Leaflet satellite map with a short rationale. Photos with GPS get a tight 1.7 km circle. For visual guesses, the circle grows from 10 km to 30 km as the model's confidence drops.

Under the hood.

Private by design.

GPS coordinates are never sent to a vision model, and API keys stay on the server. Only non-GPS hints like capture time, camera and location tags go along with the image.

Works with any of three models.

The server can call OpenAI, Anthropic or Gemini, whichever key is configured. A free Gemini key is enough for personal use.

Recovers from bad replies.

Models are asked for coordinates first, so a cut-off answer still keeps the location. Missing coordinates trigger one retry, and quota limits show a plain message.

Tested.

Unit tests cover the radius math, metadata parsing, JSON recovery from partial replies and each provider's requests.

Tech specs.

Frontend
Vite, React, TypeScript
Map
Leaflet with Esri World Imagery tiles
Metadata
exifr, with heic2any for HEIC photos
Backend
Express locally, Vercel Functions in production
Vision models
OpenAI, Anthropic or Google Gemini
Geocoding
OpenStreetMap Nominatim
Next projectCognify
All projects