Frequently asked questions

TagSmith is a free Windows app for fixing the metadata on your photos and videos in bulk – wrong dates, missing GPS, missing keywords, and more. Instead of clicking into one file at a time, you select a batch of files and fix them all at once, with a live before/after preview so you can see exactly what will change before anything is written.

Yes. TagSmith itself is completely free – no trial limits, no feature gating, no account required. The only paid part of TagSmith is optional: TagSmith Credits (coming soon), a purchasable credit balance for AI landmark identification if you'd rather not set up your own Google API key. Everything else works fully for free, forever.

TagSmith is the app. TagSmith Credits is one optional feature inside it: a purchasable, non-expiring credit balance that lets you use AI landmark identification (see below) without supplying your own Google API key. If you already have a Google API key, or don't need AI identification at all, you'll never need to think about credits.

Windows 10 or 11 (64-bit). TagSmith is a lightweight desktop app – no GPU, no special hardware, no heavy install required.

Common photo formats (JPEG, PNG, HEIC, TIFF, GIF, BMP, WebP), most common video formats (MP4, MOV, MKV, WMV, and others), and RAW files from major camera brands (Canon CR2, Nikon NEF, Sony ARW, Adobe DNG, and more).

No. TagSmith works fully offline by default – scanning folders, editing dates, adding GPS by city search, keywords, everything. It even includes a built-in database of roughly 68,000 cities worldwide for offline location search, no internet or API key needed. The only features that need a connection are the optional AI ones, covered below – and those are off by default and reset to off every time you restart the app.

For a single file, double-click its date cell in the table, type the correct date, and hit Enter. For many files at once, select them all, choose a date source in the toolbar (a typed date, or another file's date), pick which date field to write it to, and hit Apply – a preview shows exactly what will change before you commit.

Select the whole batch, choose "Custom date/time" as the source in the toolbar, type the date you know is correct (e.g. the summer the photos were actually taken), pick a destination field, and Apply. Every selected file gets stamped with that same date in one action.

Yes. Highlight the file with the correct date and use the eyedropper tool to copy it, then select the files that need fixing and paste it as their new date. This is the fix for a classic scenario: your camera's clock reset mid-trip, but a phone photo shot at the same moment still has the right timestamp.

This is what Time Shift is for. If every photo is off by the same fixed amount – six hours behind because you forgot to update the clock after a flight, for example – select the affected files, enter the offset (positive or negative), and Apply. TagSmith shifts every related date field together, so your files stay internally consistent. Shot the same trip with a second camera set to a different timezone? Select just those files and apply a different offset.

Often, yes – if the date is embedded in the filename. Apps like WhatsApp strip out all metadata when sharing a photo, but the filename (e.g. IMG-20190823-WA0003.jpg) usually still has the date in it. The Pattern Builder lets you click to mark which part of the filename is the year, month, and day (and time, if present), and TagSmith extracts a real date from it across every matching file at once. Files that don't match the pattern are simply left untouched.

Not automatically – EXIF metadata and the file system's own "Date modified" are two separate things. If you import photos from an SD card, Windows often stamps them with today's date even though EXIF has the real capture date. TagSmith can copy a correct EXIF date onto the file system timestamp too, so Explorer, Windows Search, and any app that sorts by file date shows the right order.

Select the files, type a city name into the GPS search field, pick the matching result, and Apply. Coordinates get written straight to the files – no internet or API key required, since TagSmith includes an offline database of around 68,000 cities.

GPS coordinates tell a machine where a photo was taken; Fill Location adds the human-readable version – city, state, and country – to the file's IPTC fields at the same time, so your photo library and any app reading that metadata can show a real place name, not just latitude/longitude.

Yes – that's Reverse Geocode. Select files that already have coordinates but no location text, click Reverse Geocode, and TagSmith looks up each file's coordinates individually (so files from different cities each get the right answer) and fills in the city, state, and country fields.

Yes, with GPS Propagate. This is for the common case of shooting with both a phone (which geotags automatically) and a camera (which doesn't) on the same trip. TagSmith uses your GPS-tagged files as anchors and, based on how close each other file's timestamp is to an anchor, borrows that anchor's location for it – all within a time window you control. One click can geotag dozens of camera files from a handful of phone photos.

Select the files and use the GPS Erase toggle – it strips all GPS data from the selected files in one action. Handy before posting photos publicly if you'd rather not reveal exactly where they were taken.

Double-click a file's Keywords cell to open the keyword editor. Existing keywords show as chips – click the ✕ on one to remove it, or type new ones (comma-separated for several at once) and hit Enter to add them. New keywords merge with what's already there rather than replacing it, and duplicates are caught even if the casing doesn't match exactly (e.g. "Paris" won't be added twice just because "paris" already exists).

Yes – select the files and open the batch keyword tool from the toolbar. You can add keywords, find-and-replace or delete an existing one across the whole selection, or erase all keywords entirely. Files that don't have the keyword you're searching for are simply skipped, not touched.

Yes. TagSmith writes location and keyword data to both the modern XMP standard and the legacy IPTC-IIM format at the same time, which is what most photo management software and stock agencies actually read – so you don't need to worry about which one a given app expects.

For photos with no GPS and no clue in the filename – a landmark photo from before you had a GPS-enabled camera, say – TagSmith can look at the image itself and identify where it was likely taken, using Google's Gemini AI. It suggests a location, coordinates, and keywords, which you review and choose to keep before anything is written to your files.

Yes – the Batch Wizard runs AI identification across a whole selection of files, with a progress bar, live results as they come in, and the ability to pause and resume. Results go into a review queue, not straight to disk, so you can check everything (and remove anything that doesn't look right) before applying.

A ✓ badge means the AI's guess was cross-checked against a real, authoritative location database and refined to precise coordinates. A result without it is Gemini's own estimate – usually still close, but not independently verified. Either way, nothing is written to your files until you review and confirm it.

You have two options. You can supply your own Google API key (free to set up through Google's own developer console, and the calls run directly between your computer and Google – nothing passes through TagSmith's servers). Or, if you'd rather skip that setup, you can purchase TagSmith Credits and use a prepaid credit balance instead – no Google account or API key needed on your end. Either way, this is entirely optional; every other feature in TagSmith works without it.

Every photo identified this way is a real API call to Google's Gemini AI, and Google charges for that usage – it isn't something TagSmith can offer at no cost the way the rest of the app is, since the rest of TagSmith runs entirely on your own computer with nothing to pay for. If you supply your own Google API key, you're billed directly by Google (they offer a free usage tier for casual, low-volume use). TagSmith Credits exist as an alternative: a simple, prepaid way to cover that same real cost without setting up a Google account or API key yourself.

TagSmith Credits are a prepaid, purchasable balance – one credit is used each time you identify a landmark in a photo through TagSmith's own service instead of your own Google key. They don't expire, and they're not a subscription: buy a pack, use it whenever, at your own pace. See our Refund Policy for what happens if you change your mind after a purchase.

No, not by default. Scanning, editing dates, adding GPS by city search, keywords – all of it happens entirely on your own computer. The only exception is AI landmark identification, which is off by default and only sends the one specific photo you choose to identify (as a downsized preview, never the original file) when you explicitly turn it on. Full details are in our Privacy Policy.

Many cloud-based photo-tagging services work by uploading your images to their own servers, where full-resolution copies may sit until they're processed – and potentially longer, depending on the provider's own retention practices. TagSmith takes a different approach: only a downsized preview of the one photo you choose to identify is ever sent, never the original file, and never in bulk. Nothing is stored on our servers afterward – we don't keep a copy of your images, build a database of your photo locations, or need any account tied to your photo library. Each identification is a one-off exchange: a preview goes out, coordinates come back, and that's the end of it.

We believe a landmark lookup shouldn't require handing over your entire photo library. Identifying one landmark only takes a brief look at one photo – so that's all TagSmith ever asks for, and nothing is stored afterward.

If you use TagSmith Credits (or your own billed API key), Google's own terms state they don't use that data to train their models. If you bring your own free, unbilled API key, Google's terms carry a weaker guarantee – see our Privacy Policy for the full explanation and a link to Google's current policy.

If you purchase TagSmith Credits, we store only your email and license record so we can deliver and support your purchase. Otherwise, we store nothing at all. Your photos are never stored on any server; identification results pass through and are returned straight to your device. See our Privacy Policy for everything we do and don't collect.

No. It's a one-time purchase of a credit pack, not a recurring charge. No renewal, no monthly fee.

Reach out within 14 days and we'll make it right – see our Refund Policy for the specifics.