If you want to sell photos for stock in 2026, the hard part is no longer the shooting. Cameras are good, editing is fast, and every major agency will review uploads from anyone who passes an initial application. The hard part is everything that happens after the shutter: choosing where to submit, writing metadata that survives agency search, delivering the same files to a dozen destinations without losing a working day, and understanding what each license actually pays out. This guide walks the full pipeline, using royalty terms the agencies published themselves rather than forum estimates.
What Selling Stock Photography Actually Means
Stock is a licensing business, not a sales business. You do not sell a photo once for a fee. You upload a file, an agency lists it, and every time a customer downloads it under a subscription or credit plan you receive a percentage of what the agency collected. A single file can be licensed a hundred times over five years or zero times ever, and you cannot reliably predict which in advance.
That structure has three consequences worth internalising before you start. First, there is no upfront payment, so the economics only work if files keep earning long after the shoot. Second, income is a function of portfolio size multiplied by discoverability — a thousand files nobody can find earn roughly the same as none. Third, because per-download royalties are small, the difference between a hobby and an income is almost entirely operational: how many files you can move through the pipeline per hour, and how many agencies you can reach with each one.
This is also why stock rewards a different temperament than client photography. Client work pays for a day. Stock pays for a system.
Stock Photo Royalty Rates in 2026, Agency by Agency
Royalty terms moved noticeably in 2026, and two of the biggest changes took effect on the same day. Here is what the major agencies currently publish.
Adobe Stock
Adobe pays 33% of the attributable amount on photos, vectors and illustrations, and 35% on video. The rate is applied to the net price of the license, based on the US purchase price and after any discounts the customer received. Adobe's own worked example: a plan costing $29.99 per month for ten photo licenses produces a royalty of $0.99 per licensed photo. Every contributor starts at the same rate — Adobe rewards volume through payout floors and bonuses rather than tiered percentages. We broke the arithmetic down further in how much Adobe Stock pays per download.
Shutterstock
Shutterstock runs six earnings levels for images and six separate levels for video, ranging from 15% to 40% on standard licenses and up to 50% on extended licenses. The two tracks progress independently based on how many licenses of that type you sell during the calendar year. The detail that catches new contributors: every contributor's level resets to level 1 on January 1, so you start each year back at 15% and climb again. Steady year-round uploading matters more here than on any other agency.
Getty Images and iStock
A new iStock rate card took effect on July 1, 2026. Rates are driven by download targets: hit a higher target and you receive the corresponding higher rate on new licenses from the following month, but you can only move up during the year, and your year-end download total sets your starting rate for the next calendar year. Downloads of Signature Plus files through iStock.com and Getty Images now count as two downloads toward those targets rather than one. Getty also assigns relative values to Premium Plus Video subscription downloads based on credit value, so higher-credit downloads produce a higher price per file.
Envato Elements
Also from July 1, 2026, Envato moved to a single flat 50% author fee and removed exclusivity entirely. Authors who had been on exclusive tiers keeping 75% or more saw their share cut; non-exclusive authors who kept 45% moved up slightly to 50%. The change compresses the gap between established and new contributors, and it removes the last strong argument for locking a portfolio to one platform.
The consolidation that did not happen
The proposed Getty Images and Shutterstock merger collapsed in early July 2026. For contributors the practical takeaway is that the market stays fragmented: no single destination is about to absorb the others, so the ability to distribute one file to many agencies remains the main lever on total income. For a destination-by-destination breakdown, see our comparison of where to sell stock photos.
What Sells, and What Gets Rejected
Buyers licensing stock are usually solving a specific communication problem: illustrate a blog post about remote hiring, fill a hero slot on a healthcare landing page, background a quarterly deck. That means the images that earn tend to be specific enough to match a real search and generic enough to work in several contexts. Tightly art-directed personal work often does neither.
Rejections cluster into three categories. Technical rejections come from problems invisible at screen size but obvious at 100% — luminance noise in shadows, chromatic aberration on high-contrast edges, over-sharpening halos, sensor dust in flat skies. Legal rejections come from missing model or property releases, visible trademarks, and recognisable artwork in frame. Commercial rejections mean the agency already holds thousands of near-identical files and does not need another.
The first two categories are almost entirely preventable with a review pass before submission. Our guide to Adobe Stock rejection reasons walks through each rejection code and the fix, and most of it transfers directly to other agencies since the underlying review criteria are similar.
Metadata Is the Distribution Channel
Once a file is accepted, it lives or dies by text. A buyer types a query into an agency search box and the agency matches that query against your title, description and keywords. The image itself contributes far less to ranking than contributors expect. This is the single largest gap between contributors who earn and contributors who do not, and it is entirely within your control.
Good stock metadata has a predictable shape. The title describes the scene in plain, searchable language rather than poetically. The description adds context a buyer might search for — setting, activity, mood, use case. Keywords move from most specific to most general, cover synonyms a buyer might actually type, and include the conceptual terms (teamwork, sustainability, burnout) that drive a surprising share of commercial searches.
The complication is that no two agencies want the same thing. Keyword limits differ. Category taxonomies differ. Some agencies weight keyword order, others do not. Several maintain banned-term lists and reject submissions containing them, and some require specific disclosures for AI-generated work. A single metadata set copy-pasted to every destination is either too long for one agency, non-compliant for another, or too generic for all of them.
This is the problem Rastock's metadata engine was built for. It reads the image, generates title, description and keywords, then adapts that output to each destination's rules — respecting per-agency keyword limits, applying banned and mandatory keyword lists, and mapping categories per platform. Crucially it does this with review and control rather than blind automation: you can inspect and edit before anything ships, and roll back a batch if a rule was wrong. Metadata is also written to IPTC and XMP and exportable as CSV, so the output is yours whether or not you keep using the tool.
Building a Multi-Agency Upload Workflow
Manual uploading is where most contributors quietly stop. The work is not difficult, it is just relentless: export files, open each agency portal, upload, wait, fill in metadata forms or assemble a CSV in that agency's exact column order, submit, then repeat for the next destination. Do that for eight agencies and a fifty-file batch consumes an afternoon. Do it weekly and it consumes the reason you started.
A workflow that survives contact with a growing portfolio has four properties:
- One metadata pass, many destinations. You describe the file once; the system reshapes that description per agency instead of you retyping it.
- Delivery that does not require a browser. FTP and CSV delivery run in the background while you shoot or edit.
- Per-destination status you can see. Knowing which files reached which agency, and which failed, is the difference between a portfolio and a guess.
- An undo. Bad metadata applied at scale is worse than no metadata, so batch rollback is not a luxury.
Rastock combines all four in one pass: metadata generation, bulk FTP delivery to 25+ agencies, and per-destination status tracking, rather than leaving you to stitch a keyworder to a separate uploader. There is no revenue share and no claim on your files — you pay a flat subscription and keep 100% of what the agencies pay you.
A Realistic First Year Selling Photos for Stock
Months one to three are about clearing review, not earning. Apply to three or four agencies, submit small batches, and treat every rejection as free feedback on your technical standards. Fix the causes rather than resubmitting. Expect earnings close to nothing and judge yourself on acceptance rate instead.
Months four to six are about the pipeline. By now you know what gets accepted, so the constraint shifts from quality to throughput. Automate metadata and delivery here, before the portfolio gets big enough that fixing it later means reprocessing thousands of files. This is also the point to widen from four agencies to a dozen, because with automation the marginal cost of an extra destination is near zero.
Months seven to twelve are about reading the data. Agency dashboards show which files license and which sit idle. Shoot more of what moves, drop what does not, and revisit metadata on files that get impressions but no downloads — that pattern usually means the keywords are attracting the wrong query rather than the image being weak.
None of this makes stock a fast income. It makes it a compounding one. The contributors still earning in year five are almost never the best photographers in the pool — they are the ones who built a pipeline early enough that adding a hundred files a month costs them an hour instead of a weekend.
Frequently asked questions
How much can you realistically earn selling photos for stock?
Earnings scale with portfolio size and discoverability, not with any single image. Per-download royalties are small — Adobe Stock's own documentation gives an example of $0.99 on a photo licensed through a $29.99 ten-image plan at the standard 33% rate. That means meaningful monthly income generally comes from thousands of well-keyworded files licensed across several agencies over years, not from a few strong shots.
Do I need to be exclusive to one agency?
Almost never, and exclusivity is becoming less rewarded. Envato Elements moved to a flat 50% author fee for everyone from July 1, 2026 and removed the exclusive/non-exclusive distinction entirely. Getty and iStock still offer exclusive agreements with higher rates, but for most contributors non-exclusive distribution across multiple agencies produces more total downloads than a single exclusive channel.
What actually determines whether my photos get found?
Metadata. Buyers search agency databases with text queries, so your title, description and keywords are the only thing standing between your file and a search result. Each agency has its own keyword limits, banned terms and category taxonomies, so the same file needs different metadata per destination. Files with thin or generic keywords sit unfound regardless of image quality.
Why do stock agencies reject technically fine photos?
Most rejections fall into three buckets: technical issues invisible at screen size (noise, chromatic aberration, over-sharpening, sensor dust), legal issues (missing model or property releases, visible trademarks, recognisable art), and commercial issues (the agency already has thousands of near-identical files). Reviewing at 100% zoom and checking releases before submission removes most of the first two categories.
How many agencies should a new contributor upload to?
Start with three to five and add more once the workflow is automated. Uploading manually to a dozen agencies is where most new contributors stall, because each one wants a different CSV shape, keyword count and category mapping. Once metadata generation and FTP delivery are automated, the marginal cost of adding an agency drops to near zero and wide distribution becomes the obvious choice.
Does Shutterstock's level reset each year hurt earnings?
It changes the shape of the year. Shutterstock runs six earnings levels each for images and video, ranging from 15% to 40% for standard licenses, and all contributors reset to level 1 on January 1. In practice that means January royalties are the lowest of the year and your rate climbs as licenses accumulate — an argument for keeping upload volume steady rather than uploading in occasional bursts.