---
type: concept
topic: ue-mcp-strategy
status: active
created: 2026-06-16
updated: 2026-06-16
source_notes:
  - "[[research/ue-mcp-strategy/sources/epic-monetization-model]]"
derived_from:
  - web-research
confidence: medium
shared_concept:
---
# Epic / UE Value-Capture Map

## The Money Map (plain terms, with numbers)
Epic captures value through five channels, and they are not the same size:
- **UE royalty** — 5% on lifetime revenue above $1M (3.5% if you ship on EGS at/before other stores). Engine revenue ~**$275M (2023)**. The literal "engine" channel.
- **UE enterprise seats** — $1,850/seat/yr for non-game industries (film, AEC, medical), live since UE 5.4 in 2024. AI-content-*inelastic*: you pay per human, not per asset.
- **Fortnite direct** — V-Bucks / cosmetics / Battle Pass. Not disclosed since 2022; estimated **~$3.5B-$6.2B (2023)**, ~80% of Epic's ~$5.2B.
- **UEFN engagement payout** — Epic puts **40% of Fortnite net revenue** into a monthly pool paid to islands by engagement (retention-weighted toward real-money spenders). >$722M paid to creators cumulatively.
- **Fab** — asset marketplace, Epic takes **12%** (88% to creator). Small.
- (EOS is free; EGS is 12%; both are demand-side gravity, not the prize.)

## Does "Value at the Runtime" Hold Up?
Yes — but "the runtime" is being equivocated. **[NON-OBVIOUS]** The engine royalty (~$275M) is ~13-22x smaller than Fortnite (~$3.5-6.2B). So when the thesis says "value at the runtime, not the tooling," the runtime that actually holds the money is the **Fortnite play-runtime**, not the UE engine-runtime. The claim survives, but it points at a different building than the operator instinct assumes.

## Which Channel Captures the AI-Content Wave?
Trace cheap, abundant AI content through each channel:
- **(a) UE royalty** — more games shipped on UE. But royalty only triggers above $1M lifetime per title; a flood of cheap indie/AI games mostly dies under the threshold and pays $0. Royalty scales with *hits*, not *volume*. Weak capture.
- **(b) UEFN engagement** — more/better Fortnite islands. The 40% pool is **fixed as a share of Fortnite revenue**, so more islands *redistribute* the same pool, they don't grow it — unless better content raises total Fortnite spend/retention. Capture only if AI islands convert *paying* players. **[NON-OBVIOUS]** This is the one channel where Epic's take is structurally bounded; more supply can dilute, not enrich.
- **(c) Fab** — more assets sold at 12%. Real but tiny; and an AI-asset glut compresses asset prices, shrinking the 12% base.

**[NON-OBVIOUS] The real beneficiary is none of the three cleanly — it is Fortnite's total wallet.** Value is captured when AI content makes *Fortnite itself* stickier and more monetizable (more islands worth paying in, more retention of spenders), which flows to Epic via the 60% it keeps of its own runtime, not via royalty/Fab.

## Why It Matters for the MCP-vs-Harness Question
If capture sits in the Fortnite/UEFN play-runtime, then "runtime gravity" is really **Fortnite-ecosystem gravity**, and the MCP that matters most is a **UEFN/Verse** MCP (emit playable islands), not a general UE-editor MCP. It also relocates ownership: the decision belongs to the **Fortnite/UEFN economy team**, not the Unreal Engine licensing org. The "free upside" logic holds for a Verse/UEFN MCP; it is weaker for a general engine MCP whose output (shipped games) Epic barely taxes.

## Tensions / Contradictions
- "Open the tooling, it's free, we capture at runtime" is true for the engine but the engine is the *small* channel. The big channel (Fortnite) is a walled garden an engine-neutral MCP can't feed.
- AI abundance *dilutes* a fixed 40% pool and *compresses* Fab prices — abundance can lower per-unit capture even as it raises gravity.

## Critical Questions (where this could be wrong)
- If a non-trivial slice of cheap AI games each clears $1M, royalty volume-capture revives — is the threshold actually a dead zone or a speed bump?
- Does better island supply measurably grow total Fortnite spend, or just reshuffle the pool? If the former, (b) is strong; if the latter, the bear case bites.
- Is engine royalty ($275M) understated relative to strategic value (it seeds the whole funnel)? Revenue size may understate moat value.

## Related Concepts
- [[research/ue-mcp-strategy/concepts/ue-runtime-architecture]]
- [[research/ue-mcp-strategy/concepts/mcp-as-tooling-surface]]
- [[research/ue-mcp-strategy/concepts/harness-vs-mcp]]
- [[research/ue-mcp-strategy/concepts/ai-content-generation-frontier]]
- [[research/ue-mcp-strategy/concepts/runtime-gravity-and-lock-in]]
- [[research/ue-mcp-strategy/concepts/the-bear-case]]

## Sources
- [[research/ue-mcp-strategy/sources/epic-monetization-model]]
- https://www.unrealengine.com/license
- https://sacra.com/c/epic-games/
- https://dev.epicgames.com/documentation/en-us/fortnite/engagement-payout-in-fortnite-creative
- https://www.gamedeveloper.com/marketing/epic-to-unify-content-marketplaces-and-offer-creators-88-percent-revenue-cut