Science and lab work is a permissions category before it is a metadata category. Almost every real laboratory is identifiable private property with its own photography policy, and Adobe Stock requires a property release for recognisable private interiors. Xpiks, StockSubmitter, Wirestock, PhotoTag.ai and Rastock AI all handle the metadata and delivery half competently. None of them tells you whether the frames you just shot are licensable at all, and that is where this category is actually lost.
Why this category is a permissions problem first
Adobe's property release guidance asks for a release covering any recognisable private property, artwork or location, and names identifiable exteriors or interiors of private homes and buildings alongside properties that have photography policies. A university lab, a hospital pathology suite and a corporate R and D floor all land inside that definition. The release has to be signed by the property owner, or by a corporate representative where a company owns the site, and it is what makes the content eligible for commercial licensing.
Then there are the people. A real lab has real staff in it, and anyone recognisable needs a model release before the frame can be licensed commercially. In practice this is why so much of the science imagery on the market is shot with hired models in rented or built sets rather than in working facilities: the permissions are tractable and the usage is unambiguous.
The third layer is accuracy, and it is the one unique to this category. Buyers here are medical publishers, educators, pharmaceutical marketers and journalists, and they reject frames that are technically fine but scientifically wrong: gloves missing where they would be mandatory, eyewear pushed up, a centrifuge loaded unbalanced, a pipette held like a pen. No metadata tool can see any of that. It is caught on set or not at all.
The tools, and where each one stops
Pricing in this corner of the market changes often and most vendors publish only current figures, so what follows describes each tool's model rather than a number. Check the vendor's own pricing page before committing, and treat any third-party listing of their prices as out of date by default.
Xpiks
- Best for: contributors who want a desktop application they fully control, with metadata editing and submission in one window and no cloud involved.
- Pricing model: a paid desktop licence rather than a per-file or revenue-share arrangement, which makes the cost predictable at any volume.
- Main limitation: you supply the judgement. For science content that means you type the scientific terminology, the technique names and the caption facts yourself, every time.
StockSubmitter
- Best for: breadth of destinations. Its reach across agencies is the reason it keeps coming up in contributor discussions, and for anyone submitting to a long tail of smaller platforms that matters more than polish.
- Pricing model: a paid tool with tiers tied to usage rather than a cut of your royalties.
- Main limitation: the interface shows its age, and the learning curve is real. It automates submission, not the decision about which collection a frame belongs in.
Wirestock
- Best for: contributors who would rather not run a workflow at all. You hand over files and the pipeline handles keywording and distribution for you, which is a genuine strength if your time is worth more than your margin.
- Pricing model: a share of royalties rather than a flat fee, with the share depending on the plan you choose.
- Main limitation: that share compounds over a portfolio's whole licensing life in a way a subscription does not. For science content there is a second consideration, which is that you are delegating caption accuracy on material where being wrong is expensive.
PhotoTag.ai
- Best for: fast tag generation when you already have a delivery route sorted. It is focused and it does the one job cleanly.
- Pricing model: credit or subscription tiers scaled to the number of images you tag.
- Main limitation: it stops at tags. There is no FTP delivery, no per-agency rule layer and no submission status, so it is one component rather than a pipeline. For lab imagery, generic visual tagging also tends to describe what a scene looks like rather than what is happening in it.
Rastock AI
- Best for: contributors who want metadata and multi-agency delivery in one place without giving up royalties. It generates titles, descriptions and keywords from the file, checks them against each agency's banned and mandatory keyword rules, and delivers over FTP or SFTP to 10+ agencies with per-batch status and rollback.
- Pricing model: a subscription, with a 14-day free trial that does not ask for a card. No revenue share, no ownership claim, and IPTC, CSV and XML export stay open.
- Main limitation: it works on metadata and delivery, not on releases or scientific review. If your frames need a property release you do not have, nothing in the pipeline will flag it, and specialist terminology is still worth a human pass before submission.
Five tools for science and lab submissions
Compares Xpiks, StockSubmitter, Wirestock, PhotoTag.ai and Rastock AI on AI metadata generation, FTP delivery, per-agency rule checking, revenue share and open metadata export. Pricing is described as a model rather than a figure, because vendors publish only current prices.
When to use which
Under a few hundred files a month going to one or two agencies, a desktop tool is proportionate and the cheapest path. Xpiks if you want a clean editor and full local control; StockSubmitter if your destination list is long and includes smaller platforms. At that volume the typing is annoying rather than expensive.
From several hundred files a month upwards, across more than two or three destinations, the per-agency rule layer is what starts costing real time: different keyword limits, different banned terms, different title conventions, and no single view of what actually landed. That is the case for an automated metadata and delivery workflow, and it is the point at which a revenue-share arrangement becomes the expensive option rather than the convenient one.
If you would rather not run a workflow at all, a managed pipeline like Wirestock is a defensible trade and nobody should pretend otherwise. Just price the share across the full life of the portfolio rather than against this month's income, and verify the current destination list and terms with the vendor directly before you commit a catalogue to it.
The restrictions that quietly reject lab frames
Adobe publishes a known restrictions list and two entries on it catch science shooters repeatedly. Content showing a red cross or red crescent on a white background is unacceptable for commercial use, and those marks sit on first aid kits, safety stations and wall signage in practically every working lab. Separately, content depicting identifiable NASA personnel is prohibited for commercial use, NASA trademarks and insignia are restricted, and generative AI assets depicting NASA names, logos or astronauts are not acceptable — which closes off a large share of the space-science prompts people reach for first.
Equipment branding is the third trap. Instruments carry maker's marks, screens carry software interfaces, and reagent bottles carry labels. Any of those can push an otherwise commercial frame into editorial or into rejection, and the fix is either a release, retouching, or framing that keeps the mark out in the first place.
Would this lab frame clear commercial review?
Six cases drawn from Adobe Stock's published property release rules and known restrictions list: recognisable private lab interiors, red cross and red crescent symbols, identifiable NASA personnel, equipment branding, generative AI space imagery and built sets with signed releases.
The part no tool covers
Every tool in this comparison starts after the shoot. The decisions that determine whether a science set earns anything are made earlier: whether you secured the site, whether the people in frame signed, whether the science on display is correct, and whether the frames belong in a commercial or editorial collection. Getting the release and categorisation step right is worth more here than any difference between the tools.
If you are generating science imagery rather than shooting it, the disclosure and labelling rules become part of the submission itself, which is a separate discipline covered in the agency rules for generative AI metadata. And because science and medical content overlap heavily at the buyer end, the healthcare category breakdown is worth reading alongside this one before you plan a shoot list.
Once the permissions are settled, the remaining work is keeping terminology consistent across a portfolio and getting it to each agency in the shape that agency wants. Rastock AI handles that half: metadata generated from the file, checked against per-agency banned and mandatory keyword rules, delivered over FTP or SFTP to 10+ agencies with per-batch status and rollback, no revenue share and no ownership claim. See the full feature list, or start the 14-day trial without a card from the pricing page.
Frequently asked questions
What is the best tool for science and lab stock photography?
There is no single answer, because the category splits by volume and by who owns the workflow. Xpiks and StockSubmitter are desktop tools you drive yourself and suit a few hundred files a month. Wirestock is a managed pipeline that takes a share of royalties. PhotoTag.ai focuses on generating tags. Rastock AI generates metadata and delivers over FTP to 10+ agencies without a revenue share. None of them decides whether your frames are legally licensable.
Do laboratory photos need a property release?
Usually yes. Adobe Stock requires a property release for any recognisable private property, artwork or location, and lists identifiable interiors of private buildings explicitly. A university, hospital or corporate lab is private property with its own photography policy, so a release signed by the property owner or a corporate representative is what makes the frames eligible for commercial licensing.
Why do lab photos get rejected for a first aid kit?
Because of the symbol on it. Adobe Stock lists content showing a red cross or red crescent on a white background as unacceptable for commercial use, and those marks appear on first aid kits, safety stations and signage in almost every real laboratory. The frame can be clean in every other respect and still fail on one prop in the background.
Can AI-generated science images be submitted to stock agencies?
On agencies that accept generative content, yes, but accuracy becomes the gate rather than aesthetics. Buyers in this category are publishers and educators who notice an impossible molecular structure or nonsense text on a vial label. Adobe also states that generative AI assets depicting NASA names, logos or astronauts are not acceptable, so space-science prompts carry their own restriction.
Should science stock frames go to the editorial or commercial collection?
Frames with identifiable researchers, visible institutional branding or equipment logos and no releases belong in editorial, where accurate captions matter more than keyword breadth. Frames shot with signed model and property releases and no visible marks can go commercial. Deciding this before you submit is what saves the shoot, and no tool makes the call for you.
Is a subscription tool worth it for a small science portfolio?
Below a few hundred files a month, a desktop tool you drive yourself is usually proportionate, and the free or one-off tiers cost little. The case for an automated metadata and delivery workflow strengthens with volume and with the number of destinations, because per-agency rules and delivery status are what consume time at scale rather than the keywording itself.