Epic published Fab's 2025 numbers on 22 January 2026. Three of them sit in one paragraph.
Creators earned more than $24 million. Live listings tripled to over 420,000. Publishers doubled to more than 20,000.
Every article I read quoted the first number. Nobody divided it by the other two.
The arithmetic takes ten seconds
Share $24 million between 20,000 publishers. It comes to about $1,200 each, for the whole year.
That is $100 a month, before tax and before currency conversion.
Now spread the same money across the library. $24 million over 420,000 listings is about $57 a listing, a year.
That library works out at 21 listings per publisher, on average.
One thing the total does not need is a haircut. Fab pays 88 per cent to the seller. The $24 million is already the creators' side of the split.
Both figures are floors, not ceilings
I want to be fair to Epic here. Both averages are gentle ones.
The earnings line says more than $24 million, so the top of it is open. The counts are end-of-year totals, and the publisher count doubled during the year.
So a seller who was there all twelve months did better than $1,200. Not ten times better. The size of the number holds.
A mean is the wrong tool and everyone knows it
Marketplace revenue is never spread evenly. A small number of listings carry most of it. The rest earn a little, or nothing.
So $1,200 is not what a normal publisher made. A normal publisher made less than that.
Epic's own page says event sales made up to half of a month's revenue on average. Income that lumpy is never spread evenly between sellers either.
Epic did not publish a median. In eight years of selling on marketplaces, I have never seen one publish a median.
The total caps the tail
A Fab seller writing as StraySpark posted a twelve-month retrospective on 17 April 2026. Their estimate: the very top products earn high six figures to low seven figures a year.
Take the low end of that, $700,000. Divide $24 million by $700,000. The platform has room for 34 such products.
Not 34 sellers. Thirty-four listings, out of 420,000.
The same post says mid-tier products that did $20,000 to $60,000 on the old Unreal marketplace often do $15,000 to $45,000 on Fab. Take the bottom of that, $15,000.
At that level, $24 million supports 1,600 listings. That is 0.38 per cent of the library.
I am not calling anyone dishonest. Those are one seller's estimates and the post says so. I am saying the platform total limits how many winners fit under it.
Three things Epic did not say
- Whether the $24 million counts only paid sales, or also free and Unreal-licensed content
- How many of the 420,000 listings are free downloads
- Whether all 20,000 publishers sold anything at all in the twelve months
Each of those moves the per-listing figure a long way. If half the library is free, $57 becomes $114. If half the publishers never sold, $1,200 becomes $2,400.
None of it is on the page. That is the part worth complaining about, not the size of the number.
Do this with your own statements
Take one platform and one calendar year. Divide your payout by your listing count. That is your version of $57.
Then sort your listings by earnings and read the middle one. That is your median. It will be lower than your mean, and it is the honest figure.
Last, count the listings that earned exactly zero. That count decides what you build next.
Why I keep four shelves
I sell on ArtStation, BlenderKit and SuperHive, plus 358 products on my own shop. Eighteen of those are Blender add-ons.
That is not a boast. It is this arithmetic applied to me.
One shelf at $57 a listing is a hobby. Four shelves at $57 a listing is a studio. The spread is the business model.
It is also why I keep my own copy of every listing, and why I read a support policy before I opt into it.
What the number is for
$24 million is a good result for a marketplace in its first full year. I am glad it exists and I want it to grow.
But a platform total tells you the size of the room. It tells you nothing about your chair.
Divide it. Then decide.