Breaking

GeoIntel, an Open-Source Tool, Estimates Photo Locations via Gemini API

July 1, 2026 at 03:13 EDT

AI Agents
Open Source
Foundation Models

GeoIntel, an open-source Python tool that estimates where a photo was taken using Google's Gemini API, has been released and is drawing attention among developers. It leverages the visual reasoning of a multimodal LLM to infer likely shooting locations from cues such as buildings, landscapes and signage in an image.

August 2025 · Open-Source OSINT Tool

GeoIntel: Guessing Where a Photo Was Taken — With No Metadata

A Python tool wired to Google's Gemini API infers the country, city and coordinates of a metadata-free image from visual clues alone — architecture, vegetation, road markings, shadows — returning a confidence level and its reasoning in JSON.

1.1k

GitHub stars gathered

150

Forks on GitHub

MIT

Open-source license

Confidence tiers: High / Medium / Low

How It Works — Image In, Location Out

Image or URL

no GPS metadata needed

→

Gemini API

+ Grounding with Google Maps

→

JSON Result

country · city · lat/long · reasoning

The Promise

Gemini's image understanding plus Maps grounding lets developers build high-accuracy geolocation tools with ease. Demo runs read architecture, vegetation, shadows and road markings — some call the results "scarily accurate."

The Caveat

Candidate locations are not evidence. LLMs can hallucinate wrong guesses, and cost and accuracy vary. The tool guarantees no accuracy and prohibits surveillance or stalking.

Intended Use & Boundaries

"Educational and research purposes only" — a case study in how general-purpose AI reaches into specialized tasks, and how easily it could be misused.

Install: pip install geointel

CLI: geointel --image file.jpg

Web: drag & drop, 3D Maps support

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.