Here's the summary options of Design Plan for the Marketplace: The core finding A true on-chain NIM-native NFT isn't possible — Nimiq's protocol only has four fixed account types (Basic, HTLC, Vesting, Staking) and no general-purpose smart-contract layer. This shaped everything below. Two complementary tracks (not either/or) Track A — Real NFT (Polygon + IPFS) - ERC-721 on OpenZeppelin, ERC-2981 royalty set per-token at mint time (goes to the individual creator) - Minting via an EIP-712 signed voucher: your backend checks ownership and signs off on what can be minted, but the user's own wallet pays its own gas and calls mint directly — no relayer, no hot wallet, matching how USDT payments already work in this app - Metadata pinned to IPFS via Pinata at mint time (solves a real problem: users can delete their library copy, which would otherwise orphan the NFT) - Non-custodial marketplace: atomic buy() splits royalty/fee/seller payment and transfers the token in one transaction, never holds funds or the NFT Track B — NIM Download-Rights Marketplace (reframed per your direction) - No ownership transfer — creator shares an asset (paying a fee to do so), buyers pay a NIM fee for repeatable download rights, split 80% creator / 20% treasury via two separately-verified NIM payments — no escrow - Anti-piracy: perceptual-hash comparison at share-time to block someone downloading an asset and re-sharing it as their own (images first; video hashing flagged as a harder follow-up) - Must be marketed as "collectibles," never "NFTs" — no blockchain enforces any of it Recommended order: Track B first — smaller, no smart contract, no new vendor, purely NIM (a stronger "native to the platform" story for the competition) — then Track A if there's appetite for the bigger Solidity lift.