Product and still life work breaks stock tooling in a specific way: you come back from one shoot with forty near-identical frames of the same object, and every agency wants them described differently enough not to read as duplicates. Xpiks, StockSubmitter, PhotoTag.ai, Wirestock and Rastock AI each solve a different part of that, and none of them solves all of it.
The frames are the easy part. A single session on a white sweep produces the hero, the three-quarter and the overhead, two crops of each, a version with props and a version without. All of it is sellable, and all of it describes the same object, which means the titles and keyword sets have to carry real differences or the batch starts competing with itself inside the agency's own search.
Why this category breaks ordinary metadata workflows
In most stock categories variety does the differentiation for you. A travel set contains different places; a lifestyle set contains different people. A product set contains one mug, photographed eleven ways. Pasting the same keyword list across all eleven is the obvious move and the one that hurts, because the agency's search then holds eleven equally-matched files and no reason to prefer any of them.
The second problem is volume with low variance. Product shooters are among the highest-volume contributors precisely because a controlled set is repeatable, so per-file metadata cost dominates everything else. Writing a thoughtful title for one file is pleasant; writing four hundred in a week is the actual job, and it is where most workflows quietly degrade into a copied template nobody re-reads.
The third is that still life leans heavily on conceptual keywords. A mug on linen in soft light sells as morning, calm, homeware and minimal as often as it sells as mug. Those are the terms a generic image recogniser misses, because they are not objects in the frame. Keyword strategy for this category is as much about what the picture is for as what it contains.
What to require from a tool here
Four requirements separate tools that help from tools that only type faster. First, per-file variation: the output should be genuinely different titles and keyword orders for near-identical frames, not one set stamped across a batch. Second, per-agency rules: keyword caps and banned vocabulary differ enough between agencies that a single metadata version cannot be correct everywhere.
Third, delivery that matches the volume. Browser uploads are tolerable at twenty files and painful at four hundred, so FTP or SFTP delivery with per-destination status matters more here than in low-volume categories. Fourth, no claim on the work: a controlled studio set is an asset you will re-license and re-edit for years, and a revenue share or ownership clause on it is a very different kind of cost from a subscription.
The tools, and what each one is actually for
Xpiks
Xpiks describes itself as a microstock keywording and uploading tool. It edits XMP, IPTC and EXIF metadata, does keyword autocomplete and spellcheck, supports presets and batch editing, exports CSV, and runs a pre-upload requirements check. Uploads go out over FTP, SFTP or FTPS, or through its own cloud option, and it handles photos, vectors and video. AI keywording sits on the paid tier, and the site does not print a price on the page itself. It is a desktop tool you drive: strong on precision and presets, which suits a repeatable studio set, and it expects you to already know what the keywords should be.
StockSubmitter
StockSubmitter belongs to the same desktop-submitter category: a Windows application that pushes finished files and their metadata to a long list of agencies from one queue. Its strength is breadth of destinations and the fact that it is contributor-driven rather than managed, so nothing takes a cut of the licence. Its weakness for this category is that it is a distribution tool rather than a metadata-authoring one. It moves what you wrote; it does not help you write eleven different versions of it.
PhotoTag.ai
PhotoTag.ai generates titles, descriptions and keywords for photos and video from the image itself, with a Lightroom Classic plug-in and an API for workflows that already exist. It does not deliver to agencies: the output is exported with the files or integrated elsewhere, and its own site does not publish prices on the landing page. For a product shooter who already has a working upload path and only wants the writing automated, that narrowness is the appeal. If delivery is the bottleneck, it does not touch it.
Wirestock
Wirestock is a managed pipeline rather than a tool you operate. You upload work, it handles distribution and licensing, and it pays a commission when a file licenses. Its own site presents the model as uploading once and earning commission whenever the work is licensed, pays monthly, and does not state the split on that page. For a shooter who hates the admin, that trade can be worth it. For a studio set that will earn across years, a share of every future licence is a larger number than it looks on the first monthly statement.
The agencies' own uploaders
Every major agency ships its own uploader and metadata form, and for a single-destination contributor they are adequate and free. They stop being adequate the moment the same file goes to several agencies, because each form imposes its own keyword cap, vocabulary and CSV shape, and none of them knows what the others accepted. At that point you are the integration layer.
Rastock AI
Rastock AI writes the metadata and delivers it: titles, descriptions and keywords generated from the image, checked against each destination's rules including banned and mandatory terms, then pushed over FTP or SFTP to 10+ agencies with per-destination status and rollback if a batch goes wrong. There is no revenue share, no ownership claim, and IPTC and CSV export stay open, so the metadata remains yours whether or not you keep subscribing. The trial runs 14 days with no card. It is built for the multi-agency case. If you publish to one agency only, that agency's own form is simpler and costs nothing.
Product and still life workflow tools compared
Six tool categories against what a high-volume product set needs: metadata written from the image, per-agency rule checks, FTP or SFTP delivery, video support and whether anything takes a share of the licence. Cells read unknown where the vendor's own page did not say.
When to use which
If metadata authoring is your bottleneck and delivery already works, PhotoTag.ai or the paid keywording tier in Xpiks is the smaller purchase. If delivery is the bottleneck and you prefer to write your own metadata, a desktop submitter such as StockSubmitter, or the upload side of Xpiks, solves it without touching your royalties.
If you want the whole problem to go away and are willing to pay for it out of future licences rather than up front, a managed pipeline such as Wirestock is the honest fit, as long as you price it across the life of the portfolio and not the first month. If both halves are slow and the set is going to several agencies, a tool that writes and delivers under one set of rules removes the step where a correct 49-keyword Adobe list arrives at Vecteezy well over its cap.
And if you publish to exactly one agency, use its own uploader. Every tool in this comparison starts earning its keep on the second destination, not the first.
The duplicate question no tool answers for you
Nothing in this comparison will decide how many frames of one product should go up. That is an editorial judgement. Agencies accept variants, but a set of eleven near-identical mugs with eleven near-identical keyword sets competes with itself and can read as padding to a reviewer. The usual discipline is to submit the frames that serve genuinely different uses, such as one clean hero for compositing, one styled context shot and one overhead for layout, and hold the rest back.
Where tools do help is making the submitted variants legibly different from one another, and keeping the same file from reaching the same agency twice when a batch is re-run after a failure. Both are worth testing on a small set before you trust any pipeline with a four-hundred-file week.
What it costs in time, and when a tool pays for itself
The arithmetic is unglamorous and it decides the whole question. Take the number of files you ship in a month, the minutes you spend writing metadata and uploading per file, and what an hour of your time is worth, then compare it with what a tool costs. For product and still life the file count is usually high and the per-file minutes are almost always underestimated, which is why the break-even arrives earlier in this category than in most others.
Work out your own break-even
Enter the files you ship per month, the minutes you spend on metadata and uploading per file, your hourly rate and a tool cost to see where automation starts paying for itself. All four inputs are your own figures, not measurements.
Where this leaves a product shooter
The right tool depends on which half of the work is costing you: metadata authoring, delivery, or both, and whether you would rather pay with a subscription or with a share of every licence the work ever earns. For a controlled studio set with a long shelf life, the subscription tends to look better the longer you hold the portfolio.
Whichever you pick, the requirement that does not change is per-agency correctness. A product batch is the easiest kind of work to automate badly, precisely because the frames are similar enough that one wrong template propagates across four hundred files before anyone notices it happened.
Frequently asked questions
What is the hardest part of keywording product and still life stock?
Differentiating near-identical frames. One product session yields many variants of the same object, and copying a single keyword set across all of them leaves the agency's search with several equally-matched files and no reason to rank any of them. Each submitted variant needs a title and keyword order that reflects what that particular frame is useful for.
Do I need a tool that uploads, or just one that writes metadata?
It depends which half is slow. PhotoTag.ai writes metadata and exports it without delivering to agencies. Desktop submitters such as StockSubmitter move files and metadata you have already written. Tools that do both, including Xpiks and Rastock AI, start earning their cost once you publish to more than one agency.
Is a revenue-share service worth it for studio product work?
Weigh it over the portfolio's life rather than the first month. A managed pipeline removes the admin and charges a commission on each licence, which is attractive at low volume. A controlled product set is re-licensed for years, so a share of every future licence usually costs more in total than a subscription with no claim on the work.
How many variants of one product should I submit?
No agency publishes a number, and no tool decides it for you. The practical test is whether each frame serves a different buyer use: a clean hero for compositing, a styled context shot, an overhead for layout. Variants that differ only by a few degrees of rotation tend to compete with each other rather than add reach.
Do generic AI taggers work for still life?
Only partly. Object recognition handles what is physically in the frame, such as mug, linen and wood, but still life sells heavily on conceptual terms such as calm, morning or minimal. Those are not objects and are routinely missed. Tools built for stock contributors tend to carry conceptual vocabulary; general-purpose image taggers usually do not.
Which agencies accept product shots delivered over FTP?
Most of the large ones, including Adobe Stock, Shutterstock, Alamy, Depositphotos, 123RF, Pond5 and Vecteezy, accept FTP or SFTP delivery for photo content, and Rastock AI delivers to 10+ agencies this way. Each still applies its own keyword cap and vocabulary, so one shared metadata set is not correct everywhere.