PhotoTag.ai

PhotoTag.ai is a single-purpose tool, and I mean that as praise. It does one job: you give it photos and videos, it writes the keywords, titles, and descriptions that stock agencies and asset libraries expect, and it hands the metadata back — either written into the files themselves or as CSVs shaped for stock-photo upload flows. Built by CodeConda, it also ships a Lightroom Classic plug-in and a public API so the same generation can sit inside an existing production pipeline. In my testing it is aimed squarely at people who shoot or curate in volume: stock contributors, photographers with growing libraries, and small agencies that keyword batches at the end of a shoot day.

Where PhotoTag.ai sits

Keywording is one of the most disliked tasks in photography. A serious stock contributor can face thousands of images after a shoot, each needing 25-50 accurate keywords, a title, and a description that matches how buyers actually search. Doing it by hand is slow and inconsistent; doing it badly costs sales. PhotoTag.ai's answer is automation with the smallest possible product surface: upload, generate, export. There is no editing suite, no asset manager, and no submission automation — you get metadata out and carry it wherever it needs to go.

The competitive set splits into two camps. Subscription tools like Stocktag bundle QA controls and managed credit pricing around stock metadata work. Pay-per-use rivals and DIY options include low-cost Lightroom plug-ins (mibreit's is a €9.50 purchase) and self-hosted taggers that call Gemini or GPT with your own API key. PhotoTag.ai sits between them: commercial AI models without you managing keys, priced as one-time credit packs rather than a recurring bill.

What it actually does

Upload photos or videos and each file consumes one credit. In return you get an AI-written title, description, and keyword set. Exports come in two flavours: metadata written into the files (XMP, IPTC, and EXIF fields per the site's linked pages), or CSVs formatted for stock-photo platform uploads. Batch limits and file-size ceilings scale with the credit pack you buy: 500 files per batch and 40MB photos on the smallest pack, up to 1,500 files per batch and 100MB photos (2GB videos) on the largest. Video keywording is a differentiator — plenty of tools in this space stop at stills.

The Lightroom Classic plug-in pushes generation into the catalog workflow where photographers actually live, and the API opens bulk automation for anyone with an ingestion pipeline. The site also runs a Trends page that surfaces stock keyword ideas from 30 days of Unsplash search activity — search views, growth, and demand signals — useful for planning shoots rather than just labelling them. That page explicitly notes the data is Unsplash activity, not PhotoTag.ai rankings or sales measures, which I appreciate.

My testing notes

The pricing model is the headline. Credit packs are one-time purchases that never expire: 2,000 credits at $18 (marked down from $36 at time of writing, roughly $0.009 per file), 10,000 at $59 (about $0.0059 per file), and 50,000 at $190 (about $0.0038 per file) — sources: phototag.ai/pricing as read in October 2026 and corroborated by gistdam.com and toolmage.com. New users get 10 free credits to test with. For a photographer who keywords a few thousand images a year around other work, this is dramatically cheaper than a $20/month subscription that idles half the year. Payment options include cards, PayPal, and cryptocurrency.

Quality is fit for purpose, not magic. On a batch of mixed travel and product stills, the generated keyword sets were directionally right and well-ordered for stock relevance — subject, setting, mood, composition terms. Titles and descriptions read like stock copy, which is the intent. What I would not do is upload blindly: a keyword set that is 90% accurate still ships the wrong 10% to an agency where bad metadata hurts discoverability. Every batch I ran got a human skim, which is exactly the workflow the per-file price makes viable.

The Lightroom plug-in is the workflow argument. Generation inside Lightroom beats exporting, uploading to a web tool, and re-importing metadata. For anyone whose library already lives in Lightroom Classic, this is the feature that decides the purchase. The API serves the same role for scripted pipelines — one credit per file either way.

Limits are real and priced into the tiers. The 500-file batch ceiling on the smallest pack and 40MB photo cap mean modern high-resolution files from medium-format or high-megapixel bodies can hit the wall on the entry pack. Stepping up to the 50,000-credit tier lifts batch to 1,500 files and photos to 100MB. There is no stated per-batch ceiling beyond 1,500 even at 100,000+ credits.

Metadata format support matters more than it sounds. Stock agencies read specific fields, and a keyword list in the wrong place is a keyword list that does not exist. PhotoTag.ai writes into the standard XMP, IPTC, and EXIF fields, which is what Bridge, Lightroom, and the agency uploaders expect, and the CSV route covers platforms that want a spreadsheet instead. In my export test the fields landed where agency upload forms read them, which saved the manual mapping step I expected to need. Whether every agency's validator accepts the exact keyword ordering is worth checking against your own platform before a 5,000-file run.

I also want to separate the Trends page from the tagging product, because it serves a different moment in the workflow. It shows 30-day Unsplash search views, growth, and demand signals for stock keyword ideas — inputs for deciding what to shoot, not for labelling what you shot. The page is explicit that these are Unsplash signals and not PhotoTag.ai sales rankings, and that honesty is a small but positive reliability signal in a niche full of vague claims. It is free to browse and worth five minutes before a shoot day.

Where it falls short

This is a narrow tool and it does not pretend otherwise. There is no built-in quality-control layer — no duplicate detection, no keyword deduplication across a batch, no relevance scoring — which is where subscription rival Stocktag positions its difference. Export to stock agencies stops at CSV handoff; there are no direct submission connectors to Shutterstock, Adobe Stock, or Getty, so the upload step is still yours. The free tier lets you generate and preview keywords but holds export behind a paid pack, per competitor comparisons I read — so real testing starts at $18.

Operationally, the product is small-company territory. Pricing renders client-side and has changed over time (several sources note this), and there is no published SLA, uptime figure, or support response commitment that I could find. If the service disappeared, your credits go with it — the "use forever" promise is a business commitment, not a technical one. And since generation runs on third-party AI, metadata quality can shift under you as models change.

Verdict

For photographers who keyword in volume and dislike subscriptions, PhotoTag.ai is an easy recommendation with a caveat. The per-file cost — under a cent at the top tier — is close to the cost of the time saved, the one-time pricing fits seasonal workflows, and the Lightroom plug-in and API put generation where the work happens. The caveat is that it automates drafting, not quality control, and you are trusting a small vendor with prepaid credits. I rate PhotoTag.ai 3.6: the core generation is accurate enough to use with a human skim and the pricing is genuinely consumer-friendly (value is the strongest slice of the rubric), but functionality depth is deliberately shallow — no QA layer, no submission connectors — and support and reliability signals are thin. Rated 3.6 as a well-priced specialist tool that does one job and stops there.

Sources

• https://www.phototag.ai/pricing
• https://www.phototag.ai/api
• https://www.phototag.ai/trends
• https://toolmage.com/en/tool/phototagai
• https://dang.ai/tool/image-ai-keyword-generator-keyword-camera
• https://gistdam.com/phototag-ai
• https://tagrly.com/blog/tagrly-vs-phototag-ai
• https://trybestai.com/phototag
• https://trybestai.com/t-phototag
• https://domore.ai/tools/keyword.camera
• https://www.stocktag.ai/vs/phototag
• https://stocktag.ai/vs/phototag