Fab launched in October 2024 as the unified replacement for the Unreal Marketplace, Quixel Bridge, Sketchfab's commercial offering, and what remained of ArtStation Marketplace. It consolidated four catalogs, four seller tools, and four payout systems into a single storefront. A year and a half in, we have enough data to talk honestly about what Fab delivers and what it does not.
This is a retrospective from the seller side. It covers what the first 12-18 months on Fab has actually looked like for a small studio publishing plugins and tools, what the revenue share really works out to in practice, how discoverability compares to the pre-merger Unreal Marketplace, the review and moderation realities, and the advice we would give a new seller entering in 2026. It is written for studios considering publishing on Fab, and for existing sellers who want a check on whether their experience matches what others are seeing.
The Shape of Fab in 2026
Fab at launch carried over the catalogs from its predecessor platforms — approximately 34,000 products from the Unreal Marketplace, Quixel's asset library, a substantial portion of Sketchfab's store, and a smaller transfer from ArtStation Marketplace. As of spring 2026, the catalog has grown past 180,000 products, with approximately 60% of those being newly published-to-Fab content rather than migrated inventory.
The storefront is a web app plus engine-integrated panels for Unreal Engine (the Fab plugin, mandatory in 5.6+) and native integrations for Blender, Maya, Houdini, and Godot (Godot integration launched in late 2025). Users browse, buy, and import through the native integration inside their DCC of choice. Web purchases are tied to the Epic Account, which also holds their Fortnite, Rocket League, and Unreal projects — the unified account has real consequences for discoverability, which we will return to.
Product categories are broader than the old Unreal Marketplace by design. Fab accepts 2D assets, stock footage, 3D scans, motion capture data, audio, VFX, complete games (as source code), and — new in 2025 — documented tools and plugins for DCCs other than Unreal. The broader scope means more competition in the generic categories and different dynamics for specialist products.
The seller portal has improved significantly since launch. The first six months were rough — missing analytics, delayed payouts, inconsistent tax documentation — but by mid-2025, the core seller tooling was approaching parity with what serious sellers would expect. In 2026, it is competent. Not great, not industry-leading, but workable.
Revenue Share Mechanics, Actually
Fab's headline revenue share is 88/12 in the seller's favor, which was a notable improvement over the old Unreal Marketplace's 88/12 split (same number, different qualification). The clarification matters because the old Unreal Marketplace had a 70/30 base that moved to 88/12 only after a migration period. Fab is 88/12 at base for most sellers, most of the time.
The caveats matter. Epic takes platform fees before the 88/12 split for certain transaction paths. Credit card processing fees are deducted before the split on some payment methods. International transactions incur currency conversion fees on a per-region basis. The actual money that hits a seller's bank account from a $49.99 sale after all fees, taxes, and conversions tends to land between $40.50 and $42.80 depending on the buyer's region and payment method. That is still excellent by marketplace standards, but it is not literally 88%.
Quixel's old free model shifted to a hybrid. Quixel assets migrated to Fab retained their "free for use in Unreal Engine projects" licensing but became paid for non-Unreal use. Epic pays the original Quixel contributors based on download volumes and a separate formula that is opaque to non-Quixel sellers. This matters if you are publishing scan data that competes with Quixel's library — the free-for-Unreal side of that catalog is a significant competitor that does not behave like normal inventory.
Payout timing in 2026 runs on the 15th of each month for the previous month's sales, with a 60-day hold on new sellers' first three months of revenue. The hold is aggressive but consistent with platform norms. Established sellers operate on standard 30-day terms. PayPal, direct deposit (US and EU), and Wise are supported; crypto payouts were discussed publicly by Epic in 2024 but have not materialized.
Tax documentation has stabilized. 1099-K generation for US sellers is automated, EU VAT is handled correctly through the Fab side (Epic collects and remits, sellers receive net amounts), and the platform handles most withholding tax jurisdictions through Epic's existing creator tax infrastructure. International sellers report the paperwork as significantly easier than it was on the old Unreal Marketplace.
Discoverability: The Real Story
The single most important change from the old Unreal Marketplace to Fab is how products are discovered. The old marketplace had a clear surface — featured sections, a straightforward category tree, a search that leaned heavily on exact-name matching, and a staff-pick culture that sellers could target. Fab's surface is different and, for established sellers migrating from the old marketplace, disorienting at first.
The good. Fab's search is dramatically better at semantic matching in 2026 than it was at launch. A buyer searching for "stylized medieval village modular" gets substantively relevant results even when the product names do not use those words. This helps specialist products that the old marketplace's keyword-first search would have buried. The tag system is stronger, the filtering is more granular (license type, engine version compatibility, poly count ranges), and the product page layout surfaces technical details that buyers want.
The mixed. Featured placement decisions are now algorithmic plus editorial, with the weight on the algorithm. Products that perform well in their first 30 days tend to cascade into better placement; products that launch slowly rarely recover. This is standard marketplace behavior, but it places significant pressure on launch-week marketing. Sellers who did well on the old marketplace with a slow-burn release strategy have had to adapt.
The difficult. Browse behavior on Fab is different because the storefront now serves four overlapping audiences — Unreal developers, Unity developers (Fab is accessible to Unity users, with Unity-compatible formats filtered appropriately), DCC artists using Blender or Maya, and casual hobbyists coming from the Fortnite account side. These audiences have different quality expectations, different price sensitivities, and different workflow needs. Sellers targeting a specific audience have to work harder to cut through — the old marketplace's Unreal-only audience was narrower but more predictable.
The Epic Account integration has measurable effects. Users arriving to Fab from a Fortnite context browse differently than users arriving from the Unreal Engine launcher. Analytics show that purchases from Fortnite-adjacent traffic skew toward 3D assets and visual content; purchases from Unreal-adjacent traffic skew toward tools, plugins, and gameplay systems. If your product is a tool rather than an art asset, the Unreal Engine launcher integration matters much more than the Fortnite-side surface — and that surface is behaving much like the old Unreal Marketplace did, with similar dynamics.
Search ads and promoted placement launched in mid-2025. CPMs for Fab promoted slots are lower than we expected relative to Steam ads or direct marketing, and the conversion rates for well-targeted ads to relevant product pages are reasonable. Promoted placement is a viable tool for new sellers, but it rewards sellers who have already nailed their product page and reviews before paying to send traffic.
Customer Support and Review Policies
The review system is where Fab has made both meaningful improvements and introduced new frictions.
Fab reviews now require a verified purchase. This eliminates the review-bombing behavior that occasionally plagued the old marketplace. Reviews are also tied to engine version and use case — a reviewer marks whether they used the product in Unreal 5.6, Blender 4.2, or another environment, and other buyers can filter reviews by their own context. This is genuinely useful for multi-engine products where performance varies by target.
The review-response system lets sellers respond to reviews publicly, with clear editorial boundaries. Epic moderates seller responses for professionalism but does not generally remove critical reviews. The moderation leans buyer-side — if a review is harsh but accurate, it stays up. Sellers who engaged with reviews professionally saw better retention on their product pages than sellers who ignored negative feedback.
Refunds are handled on Epic's side through the unified account refund system. Sellers see refund volume in their portal but do not directly approve or deny refunds. Refund rates in 2026 average around 3-5% of sales, skewing higher for complex tools and lower for art assets. Refunds during the first 14 days are nearly automatic for standard reasons (compatibility, missing features, buyer error); refunds after 14 days require case-by-case review. Sellers of complex tools report that the refund process is reasonable for genuine issues but occasionally allows bad-faith refunds on high-price products — Epic has iterated on this and the situation in 2026 is better than it was at launch.
Customer support for buyers is acceptable. Support for sellers is better than it was on the old Unreal Marketplace, which had a reputation for slow and opaque seller-side responses. Fab's seller support uses a dedicated portal with reasonable response times (we see 24-48 hours typical, longer for complex cases). Dedicated account management exists for high-volume sellers but is not widely advertised — crossing a revenue threshold typically results in Epic reaching out directly rather than sellers requesting it.
Content Moderation and Policy
Fab's content moderation in 2026 is stricter than the old Unreal Marketplace and noticeably stricter than what Sketchfab allowed pre-merger. This matters if your product sits near any line — AI-generated content, content scraped from other sources, adult or violent content, brand-adjacent content.
AI-generated content policy has evolved. Pure AI-generated assets (generated from prompts, no significant human authorship) were banned at launch and remain banned. Hybrid workflows — human-authored assets with AI-assisted steps like texture generation, upscaling, or procedural variation — are allowed if disclosed. The disclosure is checked on submission and spot-checked through user reports. Sellers using AI in any part of their pipeline should disclose it honestly. Undisclosed AI content, when caught, results in product delisting and potential account action.
Brand and IP moderation is aggressive. Products that reference trademarked brands, IP, or recognizable real-world products (even in descriptions) are reviewed carefully. The phrase "inspired by" does not help. The moderation is not arbitrary — it follows Epic's broader IP policy, which is conservative — but it catches sellers by surprise when their "sci-fi helmet inspired by [game]" listing is rejected.
Derivative content policy is clearer than on the old marketplace. Fan content, unofficial implementations of proprietary systems, and content that could be confused with official products is rejected consistently. This is good for buyers and good for the marketplace's long-term health, but it rejects a category of products that occasionally slipped through on the old system.
Who Is Actually Making Money on Fab in 2026
Revenue distribution on Fab is heavily skewed, but the shape of the distribution is different from the old Unreal Marketplace. Specifically:
Top-tier sellers (top 200 products by revenue) earn more than their equivalents did on the old marketplace, because Fab's larger and more diverse audience expands the addressable market. The very top products on Fab earn in the high six figures to low seven figures annually — numbers that were unusual on the old Unreal Marketplace.
The middle tier has compressed. Products that would have done $20,000-60,000 annually on the old marketplace often do $15,000-45,000 on Fab. The increased competition and the broader catalog have made the middle harder. Established sellers have maintained revenue mostly by expanding their catalog — they ship more products now than they did pre-merger, and the aggregate revenue has held.
Long-tail sellers have benefited. The combination of better semantic search, multi-engine compatibility, and a broader audience means that a well-made niche product finds its buyers more reliably on Fab than it did on the old marketplace. Several sellers we know personally who published consistently-but-quietly on the Unreal Marketplace have seen 40-80% lifetime revenue increases on Fab versus projections from their old-marketplace data.
New sellers (first year on Fab, no prior marketplace history) have a steeper initial curve than new sellers did on the old Unreal Marketplace in its mid-life. The catalog is larger and more competitive. The first three to six months of a new seller's Fab presence typically generates less revenue than the equivalent period would have on the 2022-era Unreal Marketplace. However, the ceiling is higher, and sellers who clear the initial curve often see stronger subsequent growth than they would have on the predecessor platform.
Advice for New Sellers in 2026
Based on observing what has worked and what has not for sellers entering Fab over the last year:
Pick a category where you have genuine expertise. The broader catalog means that generic quality is everywhere. A product that demonstrates specialized knowledge — technical depth, a specific workflow solved well, a category the platform underserves — has a much better discoverability path than a product that is merely competent.
Invest in the product page. Fab's product pages support detailed technical documentation, embedded video, interactive 3D previews (for art assets), and structured compatibility data. Buyers in 2026 use these features heavily. A thorough product page with demonstrable information outperforms a sparse one even when the underlying product is identical.
Launch with a content plan, not just a product. Release with two or three follow-up updates planned for the first 90 days. Product updates push the product back into buyers' feeds and into the "recently updated" discovery surface. Sellers who release once and then go silent almost universally underperform their own product's potential.
Use the multi-engine compatibility surface honestly. Fab's filtering lets buyers find products for specific engines. Mislabeling a Unreal-only asset as Unity-compatible gets refunds, bad reviews, and eventual delisting. Products that genuinely support multiple engines — and clearly document how — capture the cross-engine audience that Fab's larger catalog reaches.
Price with the category, not against it. Fab's pricing transparency and the presence of many similar products in each category mean that the price anchors of each niche are visible. Pricing significantly above the category anchor requires demonstrable differentiation; pricing significantly below often signals (accurately) that the product is lower quality. Middle-of-category pricing with differentiated positioning is the steady path.
Build a direct audience outside Fab. The sellers who weather marketplace changes are the ones who can reach their buyers without relying on marketplace discoverability. A newsletter, a Discord, a portfolio site, a YouTube channel — whatever fits your work — gives you a moat against Epic's algorithm changes, which do happen. Sellers whose entire audience lives on Fab are captive to platform decisions.
Do not rely on free assets for the funnel. The free asset program exists on Fab (the monthly featured free assets continue, with a different model from the Unreal Marketplace's free-of-the-month program). Giving away a free asset to drive sales of paid products has a smaller effect on Fab than it had on the old marketplace. The conversion paths are less direct.
For sellers of tools and plugins specifically — not art assets — the dynamics favor technical depth and good documentation. Our own experience publishing tools like the Unreal MCP Server suggests that tools which solve specific, well-articulated problems (automated asset operations, batch workflows, integrations with external systems) retain their buyers better than general-purpose offerings. The tool market on Fab is less crowded than the art-asset market, and the buyers are more discerning about ongoing support.
What We Got Wrong Predicting Fab
A few of our own early predictions from the launch period have not held up, and it is worth naming them.
We expected Fab to hurt smaller sellers more than it has. The first six months were rough for everyone, but by year one, the combination of better search, broader audience, and catalog growth had lifted most capable sellers back to or past their pre-merger revenue. The smallest sellers — those with a single low-quality product — did struggle. Competent sellers mostly recovered.
We expected Epic to monetize more aggressively than they have. Promoted placement is cheaper than we expected, the featured slots are still substantially editorial, and the pressure on sellers to buy ads to remain visible has not materialized in the way we worried about. Epic has been patient with the platform's growth, and the monetization levers that exist are used lightly. This may change, but a year and a half in, the platform is not hostile to sellers.
We expected the multi-engine expansion to matter less than it has. The Blender, Maya, and Houdini integrations have added significant audience, and the Godot integration has been a pleasant surprise — the Godot audience is small but engaged, and early-to-market Godot content has outperformed our modeling. Sellers who treated Fab as Unreal-Marketplace-with-a-new-name have left money on the table; sellers who embraced the multi-engine scope have captured it.
The Bottom Line
Fab in 2026 is a real marketplace, doing the things a marketplace does. It is not the easy-money environment some sellers hoped for, and it is not the seller-hostile environment others feared. It is competitive, it rewards quality and consistency, and the mechanics are stable enough in 2026 that a seller can build a business on it without constant platform anxiety.
For sellers considering entering, the answer is: yes, but go in with realistic expectations. The first year will be slower than you want. The second year, if your products are genuinely good and you keep shipping, will be substantially better. The fifth year is where the compounding of catalog size, reviews, and platform reputation becomes obvious — and the sellers who started three or four years ago on the Unreal Marketplace and transitioned to Fab are the ones seeing that compounding now.
For existing sellers, the advice is simpler. Fab has settled. The tooling works. The audience is larger than the old marketplace's. The competition is stiffer. Keep shipping, keep updating, and keep the quality bar high. The platform rewards that pattern, and it has done so consistently for the last twelve months.