Best Zero Footprint DICOM Viewer for Multi-Site Imaging Groups

The short answer: for a multi-site imaging group, the best zero footprint DICOM viewer is the one that is cleared for the job each group of users actually does, priced in a way that survives adding a site, and hosted where your security review can live with it. Zero footprint only describes deployment: the viewer runs in a browser with nothing installed. It says nothing about whether the product is cleared for diagnostic reading, and vendors sell both kinds under names that look identical on a datasheet. eUnity, Ambra ProViewer and the Merge Universal Viewer each hold their own FDA 510(k) clearances. Plenty of browser viewers do not. Ask for the K number before you ask for the price.

Updated August 2026. Clearance records in this article were pulled from the openFDA 510(k) API on August 31, 2026 and can be checked by K number at accessdata.fda.gov.

What zero footprint actually means, and what it does not

A zero footprint viewer is a web application. The images render in the browser, nothing gets installed on the endpoint, and there is no client to patch on 400 referring physicians' laptops. That is the entire promise, and for a group running four imaging centers and a hospital contract it is a real one. The alternative is buying PACS seats for people who look at two studies a month.

What the term does not tell you is any of the following: whether the viewer is cleared for primary interpretation, whether the images stay in your archive or get copied to the vendor's cloud, whether it works on the browsers your hospital partners actually allow, or what happens to the price when site five opens. Those four questions decide the outcome of the purchase. The deployment model does not.

The clearance point is worth being blunt about, because it is the one that gets skipped. A viewer marketed for diagnostic reading in the United States is regulated under 21 CFR 892.2050 as a medical image management and processing system, a Class II device, usually cleared under product code LLZ. A viewer marketed for reference or review is not. Both can be zero footprint. We work through the whole tier structure, with verified K numbers for each tier, in DICOM viewer software and which DICOM image viewers are FDA cleared.

How crowded this category actually is

People assume a browser viewer with a clearance is exotic. It is not. Product code LLZ has produced a steady stream of clearances every single year, and 85 of the devices cleared under it have the word viewer in their FDA device name.

Source: openFDA 510(k) API, product code LLZ, queried August 31, 2026. That is 471 clearances in under seven years, spread across at least 100 different companies. The category is not hard to enter and it is not thinly regulated.

Why that matters in a negotiation: if a vendor tells you clearance is unnecessary for their product, they are describing a business decision, not a regulatory obstacle. Sixty to eighty companies a year manage it. Ask what the product is marketed for instead, get the answer in writing, and make sure it matches what your radiologists will do with it.

What is the best zero footprint DICOM viewer?

There is no single winner, and any article that hands you a ranked list of five is guessing on your behalf. The honest structure is that the answer changes with who the users are. Diagnostic reading needs a cleared viewer on a calibrated display. Referring physician access needs breadth, single sign-on and a support path that does not run through your radiologists. Outside sites and patients need image exchange more than they need a viewer at all. A product that is excellent at one of those is frequently mediocre at the next.

So the useful exercise is to count your user populations before you take a demo, and to price each one separately. Most multi-site groups discover they have three: about ten to forty radiologists who read and sign, a few hundred referrers who look, and a long tail of outside requests that arrive by phone and CD. Only the first group needs a diagnostic license.

Five questions that separate zero footprint viewers

The third row is the one that costs real money. Viewer contracts are frequently signed by a group with three sites and renewed by a group with seven, and the growth clause is much cheaper to negotiate before signature than after. The same pattern shows up across imaging software categories, which we cover in the radiology software contract terms worth negotiating.

Named users, concurrent users, or per site

These three licensing shapes look similar in a quote and behave completely differently at four sites.

Named user pricing charges for every person with an account. It is predictable and it punishes exactly the use case a zero footprint viewer exists for, because a referring physician who logs in twice a quarter costs the same as one who logs in daily. Concurrent pricing charges for simultaneous sessions, which fits referrer access well and fits diagnostic reading badly, since your radiologists are all logged in at once during the day anyway. Per site pricing is the cleanest for a growing group until you look at what a site is defined as, and discover that the definition was written by the vendor.

The practical approach is to split the contract by population rather than trying to win one number. Buy named diagnostic licenses for the radiologists, because that count is stable and you want the calibrated, cleared path there. Buy concurrent or per site for the referrer population, because that count is not stable and you do not want it to be a budget line you manage. Groups that try to cover everyone under a single unit almost always overpay on one side of that split.

Vendor hosted or on premises

Most zero footprint viewers are now offered both ways, and the decision usually gets made by whoever is loudest in the room rather than on merit. The real trade is straightforward. Vendor hosted means the availability, patching and browser compatibility problem belongs to the vendor, and your security review has to cover a third party holding or streaming protected health information. On premises means the images never leave your network, and your IT team has just inherited a production web application to keep patched, monitored and available to a few hundred outside clinicians who will call your PACS administrator when it is slow.

That second half is genuinely underestimated. Running an internal clinical web service is closer to ordinary web operations than to running imaging hardware: it needs staged updates, a rollback path, and the ability to patch and redeploy without taking the service down in the middle of a clinic day. If your team does not already work that way, vendor hosting is usually the cheaper answer even when the license looks more expensive, and the honest version of that conversation happens before the contract rather than after the first outage.

What multi-site groups get wrong

Three mistakes recur.

The first is assuming zero footprint means diagnostic. It describes how the software is delivered, not what it is cleared to do, and buying a review viewer for a reading room is a problem that surfaces at the worst possible moment. The second is buying the viewer to solve an archive problem. If studies from four sites cannot be seen in one place, the gap is usually storage and routing rather than display, which is enterprise imaging and vendor neutral archive territory, and a viewer bolted on top will not fix it.

The third is treating the viewer as the whole answer to referring physician satisfaction. Referrers complain about images because they cannot find the report, not usually because the images render badly. Report turnaround and result delivery are a separate problem with a separate fix, and we go through what actually moves it in radiology report turnaround time and what moves it.

Where an AI assistant fits into this

Not in the viewer. Radiological.ai is decision support that connects to the PACS you already run over standard DICOM, flags suspected findings for a second look, pushes time-critical studies up the worklist and drafts the structured report into your template. It does not display studies for primary interpretation and it does not replace whatever viewer you land on. Whichever browser viewer your referrers end up using, the reading path underneath it is unchanged, and the responsible radiologist reviews, edits and signs every study.

If you are earlier in the process and still working out which layer of the imaging stack is actually failing, start with the DICOM viewer software comparison for the clearance question, then PACS software compared by study volume if the reading platform is the real problem. For the money side, radiology software pricing collects what federal contract data shows about each software category.