Is stock illustration saturated? Here are the counts
Not an opinion about the market — the actual size of the catalogue you would be uploading into, subject by subject.
Measured · SincereTrend
- 91,279
- illustrations already on Adobe Stock for the median keyword
- 48.8%
- of tracked keywords already have 100,000+ illustrations
- 20.5%
- already have more than a million
"Is it saturated?" is usually argued with anecdotes — your own downloads, someone's port review, a forum thread. The question has a factual half, though, and it is answerable: for a given subject, how much already exists? We count that every day, so here is the distribution.
How much already exists
For the median tracked subject there are 91,279 illustrations already on Adobe Stock. 48.8% of tracked subjects carry more than 100,000. 20.5% carry more than a million. At the 90th percentile the count is 2,863,266.
| Marketplace | Keywords read | Median catalogue | 90th percentile |
|---|---|---|---|
| Adobe Stock | 3,265 | 91,279 | 2,863,266 |
| iStock | 3,621 | 27,379 | 492,209 |
| Etsy (listings) | 7,023 | 12,956 | 514,045 |
What "saturated" should mean
A number like 91,279 is easy to read as despair and that is not what it supports. Saturation is not a property of a market; it is a property of a subject, and the spread here is enormous — from a few hundred assets to nearly three million. The useful question is never "is stock saturated" but "is this subject saturated, relative to how many people are asking for it".
Which is why a supply count on its own decides nothing. It is a denominator. A subject with 200,000 existing illustrations and heavy, growing demand can be a better bet than a quiet one with 2,000 and nobody looking — and the reverse is true just as often. Supply becomes information only next to demand, and next to how much of that supply is generated: we measured that separately in how much of Adobe Stock is AI, where the median subject came out at 74.7%.
The honest summary of the counts: the broad, obvious subjects are effectively closed, and they have been for years — not because of AI but because a decade of contributors got there first. What is left is narrower and requires actually knowing which narrow thing anyone wants.
Method and limits
- Sample: 3,265 keywords with an Adobe Stock reading, using the most recent unfiltered illustration count for each.
- Illustration and vector catalogues only, on every marketplace. Photography is not counted anywhere in our pipeline, so nothing here describes the photo market.
- These are marketplace-reported result counts. They are what a buyer searching that phrase would be shown, which is the number that matters competitively — but they are not an audited inventory, and they include assets that would never surface on page one.
- Etsy counts listings, not assets, and those listings come from independent sellers. It is in the table because seller competition is real, not because it is the same measurement.
- Coverage differs by marketplace, so the three rows are not a like-for-like comparison of library size. Each row is a distribution over the keywords we actually read there.
Per-subject supply, demand and AI share are free on every keyword page — browse the corpus, or read the method in what to draw next.