Does HoudahGeo have AI location detection?

No. HoudahGeo is built around GPS tracks and manual map-picking — it writes coordinates you supply to it, either from a GPX file you recorded or from a location you pin yourself. It has no computer vision, no landmark recognition, and no way to infer where a photo was taken from the image content. If you don't have a track and don't remember the location, HoudahGeo's map is an empty question: it needs an answer before it can help. That gap is precisely what RetroTagr's AI is built for.

Does RetroTagr support GPX track import?

Not currently. RetroTagr's geotagging path is AI visual recognition plus a manual map — you either let the AI suggest a location from the image content, or you pick one yourself. If you have a GPS track recorded during a shoot, a track-matching tool like HoudahGeo (Mac) or GeoSetter (Windows) will give you more precise results by timestamp than AI inference can. The two approaches cover different problems: track matching for shoots where you recorded your route, AI for archives where no track exists.

Can HoudahGeo help with old photos whose locations I don't know?

Only if you can tell it where they were taken. HoudahGeo's manual map picker is capable, but it needs you to know the location first — you navigate the map to the right spot and assign it. For family archives, inherited scans, or any photo taken before GPS was standard, the problem is often that the location isn't known at all. In those cases HoudahGeo has nothing to work from; AI inference is the only automated alternative.

If I already use HoudahGeo on a Mac, do I need RetroTagr?

It depends on whether your whole library has GPS tracks. If you consistently carry a GPS logger and your archive is well-covered, HoudahGeo handles everything on Mac. RetroTagr adds value when part of your archive predates your GPS habit — old photos you imported, inherited images, trips where you forgot to start the tracker. For those photos, AI location suggestions give you a starting point that HoudahGeo's map simply cannot provide.

RetroTagr vs HoudahGeo: AI Location vs GPX Workflow

Honest head-to-head: HoudahGeo is the polished Mac app for GPS-track photographers. RetroTagr is the AI-first web tool for photos with no track at all. When each wins.

HoudahGeo is the most polished Mac geotagging application available — actively maintained (version 7.2 shipped in June 2026), built around GPS track synchronisation, and priced at a one-time $39. RetroTagr is the web-based alternative built for a different problem: tagging photos when you don't have a GPS track and don't know — or can't remember — where they were taken. This page compares them directly so you can decide which fits your library.

TL;DR: Pick HoudahGeo if you shoot on a Mac, regularly carry a GPS tracker or a phone recording your route, and want a polished native app that syncs track to photo automatically for a one-time $39. Pick RetroTagr if you don't have a GPS track (or never will for your archive), you're not on macOS, or you need AI to suggest a location from the image content itself. Both write the same standard EXIF GPS — the choice is entirely about workflow, not lock-in.

Feature	RetroTagr	HoudahGeo
Cost	Free tier (100 photos + 5 AI); one-time photo packs from €1.99 — no subscription	~$39 one-time purchase (includes 12 months of updates)
Where it runs	Web app (any OS, any browser)	macOS only — no Windows, no Linux
Requires a location source?	No — AI infers location from image content	Yes — GPX track or manual map-pick
AI location detection	Yes — visual recognition, landmark detection, confidence score	No
GPX track sync	Not currently	Yes — GPX, NMEA, Garmin FIT, Wintec TES, and more
Best for	Locations you don't know; any platform; trackless archives	Mac photographers with GPS trackers; precise timestamp matching

HoudahGeo — fairly described

HoudahGeo is purpose-built for GPS track synchronisation, and it executes that job exceptionally well. The workflow it is designed around: you carry a GPS device — a sports watch, a dedicated GPS unit, a phone running a track-logging app — alongside your camera. When you get home, you hand HoudahGeo the GPX file and your photos. It matches each image's capture timestamp to the nearest track point and writes the corresponding coordinates into the EXIF — automatically, across the entire shoot. The result is coordinates that reflect where you were standing at the exact moment you pressed the shutter.

Strengths:

One-time purchase. At around $39 HoudahGeo is a buy-once tool. The purchase includes 12 months of updates; beyond that, update extensions are available at a discounted rate. For photographers who dislike monthly software subscriptions, that model is genuinely appealing.
GPX track sync. Precise timestamp matching to recorded tracks is the most accurate geotagging method available — coordinates derived from real GPS, not inference. HoudahGeo supports GPX, NMEA, Garmin FIT, and Wintec TES formats, so it works with virtually any GPS device or phone track-logging app.
Reverse geocoding. Turns the raw coordinates into human-readable city, region, and country names, written into the relevant IPTC and XMP fields alongside the GPS data.
Video geotagging. Handles drone footage, action camera clips, and any GPS-tagged video format — not just still photos.
Mac-native experience. Integrates naturally with Apple Photos and Lightroom; feels like a first-class Mac application because it is one.
Actively maintained. Version 7.2 shipped in June 2026 with expanded reverse geocoding and interface improvements. This is a tool with a developer behind it.

Trade-offs:

macOS only. There is no Windows version, no Linux port, and no browser interface. If you are not on a Mac, HoudahGeo is simply unavailable — the comparison ends there for you.
Requires a location source. HoudahGeo cannot analyse an image to suggest where it was taken. Its manual map picker is capable, but it needs you to navigate to the right spot — which means you need to know the location first. For photos where the location is the unknown, HoudahGeo offers no path forward.
Free trial, no permanent free tier. HoudahGeo offers a trial period before purchase, but there is no free ongoing usage once it expires. Evaluating the tool is easy; committing is a $39 decision.

HoudahGeo is the right answer if you are a Mac photographer who builds GPS logging into your routine. For that specific workflow it is hard to beat.

RetroTagr — fairly described

RetroTagr is built for the case where you do not have a location source. Import a set of photos and the AI analyses the visual content of each image — architecture, landmarks, street signage, terrain, vegetation, vehicles, coastlines — and suggests a location. Each suggestion comes with a confidence level so you can judge how much to trust it: accept the pin as placed, drag it to refine the spot, switch to manual map-picking, or skip the photo if the image gives too little to work from.

Strengths:

AI location detection. The only automated path for photos whose location you don't know. Landmark-heavy tourist shots land at street level; distinctive non-famous places typically resolve to the right town or region; generic interiors and featureless landscapes are flagged low-confidence for manual review rather than guessed wrongly.
Runs anywhere. A web app — Windows, Mac, Linux, ChromeOS, any modern browser. Nothing to install.
Free tier. First 100 photos and 5 AI suggestions cost nothing. You can test AI geotagging on your actual photos, your specific batch of unknowns, before deciding whether to pay.
Manual map too. For photos whose location you do know, RetroTagr has a map picker alongside the AI path — you are not forced into one method.
Actively developed. Receives ongoing updates and support.

Trade-offs:

Paid above the free tier. Beyond the free 100 photos and 5 AI suggestions, you buy one-time photo packs (from €1.99) — you pay for the photos you locate, with no subscription. Unlimited tagging is not free.
Cloud-based. Photos are uploaded for AI inference. They remain private to your account and are not used for model training, but if local-only processing is a hard requirement, HoudahGeo satisfies that condition and RetroTagr does not.
No GPX track sync yet. For shoots where you recorded a track, HoudahGeo's timestamp matching will give more precise coordinates than AI inference can for those specific photos.
AI accuracy is graded, not uniform. Famous landmarks resolve well; generic scenery may not resolve at all. The confidence score tells you which is which.

RetroTagr is the right answer when the location is the unknown — when the problem is not just writing a coordinate you already have, but figuring out what coordinate to write.

Decision scenarios

"I'm a Mac photographer and I carry a GPS watch on every hike." → HoudahGeo. This is precisely the workflow it was designed for: accurate timestamp sync to a recorded track, across a full shoot, in a polished Mac app.
"I have 800 old family photos, no GPS tracks, and no memory of where most were taken." → RetroTagr. HoudahGeo's map cannot fill in the blank when the location is genuinely unknown. AI inference is the only automated option for this kind of archive.
"I'm on Windows and want to geotag from GPS tracks." → Not HoudahGeo — it's Mac-only. Look at GeoSetter (Windows, free, GPX track sync) for the track-matching workflow, or use RetroTagr's manual map for photos you can place yourself.
"I want to test geotagging before buying anything." → RetroTagr's free tier covers 100 photos and 5 AI suggestions at no cost. HoudahGeo has a trial period but requires purchase to keep using it.
"Some of my photos have a GPX track and hundreds more have no location data at all." → Use both. HoudahGeo for the track-synced shoots on your Mac; RetroTagr for the trackless portion of the archive. The output of both is standard EXIF GPS, so the two workflows compose cleanly.

Migrating between them

Both tools write the standard EXIF GPS tags (GPSLatitude, GPSLongitude, GPSAltitude), so there is no proprietary format and migration is a metadata read in either direction.

HoudahGeo → RetroTagr: HoudahGeo writes GPS coordinates directly into your files. Upload the folder to RetroTagr and run its AI on the photos that still need a location.
RetroTagr → HoudahGeo: Download your tagged library from RetroTagr (the embedded EXIF option), then open the folder in HoudahGeo. Your coordinates appear on its map immediately — HoudahGeo reads the same tags it writes.

No re-tagging, no conversion step, no format lock-in on either side.

Can you use both?

On macOS, yes — and for a mixed archive it is the sensible split:

Run RetroTagr across the whole library first. Let AI suggest locations for everything, bulk-accept the high-confidence results, and flag what needs review.
Download the tagged files.
Open that folder in HoudahGeo. The photos AI already located show up placed; the ones still missing can be finished with GPX tracks (if you have them for any of those shoots) or manual map-picking.

Both tools read and write the same standard EXIF GPS tags, so neither disturbs the other's work — HoudahGeo reads whatever coordinates are already in the file. If you carry a GPS tracker going forward but have a legacy archive to clear first, starting with RetroTagr and maintaining with HoudahGeo is a reasonable long-term split.

Where to go from here

If you're deciding between AI-based and track-based geotagging more broadly, the HoudahGeo review covers when each approach fits in more depth. For the most detailed comparison against the free Windows track-matching alternative, see RetroTagr vs GeoSetter. Ready to test AI geotagging on your own photos? The free tier at retrotagr.com/#pricing covers 100 photos and 5 AI suggestions with no payment required.