A research team has built, run and formally verified a blockchain framework that converts image, video and PDF files into non-fungible tokens, and then measured how much minting actually costs on a test network. According to Bioengineer.org, whose report on the study was published on 9 November 2026, the study appears in the journal Discover Networks. The average gas consumption for minting one token was 282,108 gwei.
The source describes gas as a fee deducted during wallet approval when a token is minted. The number matters because it is one of the few hard measurements in a field where most papers stop at theory. The team — Jhuma Dutta, Nisha Yadav, Ayan Pal, Subhas Barman, Matangini Chattopadhyay and Samiran Chattopadhyay — tested files ranging from 1.92 to 427 kilobytes across PNG, GIF and PDF formats. Those sizes are small. The framework is built for documents and graphics, not feature-length video. We covered a connected angle in Blockchain technology explained: blocks, nodes and consensus.
What the framework actually does
The system pairs an Ethereum smart contract written in Solidity with a decentralised storage network called the InterPlanetary File System, or IPFS. The split is the core design decision. Storing a full image or PDF directly on a ledger would be slow and expensive. Instead, IPFS computes a unique cryptographic hash of the uploaded file, known as a Content Identifier. Only that identifier and its storage address go on-chain. The file itself lives off-chain, and the hash lets anyone later confirm the file has not been altered. For related coverage, see What is a smart contract, and how does it actually work?. For related coverage, see What is a smart contract, and how does it actually work?.
Minting works in a defined sequence. A user uploads content and sets metadata: title, description, file size and a royalty percentage for future sales. A contract inheriting from the ERC-721 token standard then generates a distinct token identifier, links it to the storage address, and increments an internal counter. After wallet approval and a gas fee, the token is registered on the ledger and filed in the owner's personal repository, which the authors call an NFT Treasury.
How trading works in the contracts
Two contract procedures govern the marketplace. makeItem handles listing: it checks the price is greater than zero, moves the token into the marketplace contract's escrow, and emits an Offered event across the network. purchaseItem handles the sale: it calculates the total cost including platform fees, verifies the payment, sends Ether to the seller, remits fees to an immutable fee account, transfers the token to the buyer, and emits a Bought event.
According to Bioengineer.org, the researchers employed Proof-of-Stake consensus mechanics alongside the ERC-721 standard. They also built on OpenZeppelin contracts version 4.9.5, community-reviewed open-source libraries whose ReentrancyGuard module defends against a known class of smart contract exploit in which an attacker re-enters a contract mid-transaction to drain funds.
Why the team says the work was needed
The authors frame the project as a response to gaps in earlier deployments. Existing token environments, they write, suffer from poor cross-platform interoperability, high energy expenditure, piracy vulnerabilities, weak ownership controls, and smart contract execution flaws. Non-fungible tokens need unique, non-interchangeable identifiers to prove scarcity and provenance — unlike fungible currencies such as Bitcoin or Ether, where any one unit equals any other.
The study also makes a blunt observation about the academic literature. Commercial platforms including OpenSea, Axie Infinity, CryptoPunks, Foundation, Rarible, AtomicMarket and SuperRare have established substantial trading volumes, but scholarly work has often lacked concrete implementation frameworks or standardisation for pricing and cross-chain scalability. Earlier studies covered economic valuations, museum fundraising and healthcare data, yet frequently omitted practical token blueprints, formal contract verification, or latency measurements across multimedia formats. This paper is an attempt to fill that hole with something reproducible.
How it was tested, and what the numbers cover
The evaluation ran on a physical test bench: an Intel Core i5-1035G1 processor at 1.00 GHz. The backend targeted the Ethereum Sepolia testnet to emulate Ethereum 2.0 network conditions, and the user interface, built with the React JavaScript library, was deployed on Vercel. MetaMask served as the wallet and payment gateway.
One measurement survives in the published summary: the average minting gas cost of 282,108 gwei. The source text cuts off before reporting the comparison figure it was set against, so any broader cost conclusion would go beyond the evidence. That is worth stating plainly. A single average on a testnet does not tell a reader what minting would cost on Ethereum mainnet, where fees move with network demand. The testnet was chosen to emulate Ethereum 2.0 conditions, not to price real transactions.
What is established, and what is not
The evidence supports this: a six-author team built an end-to-end architecture on standard components, verified it formally, and produced at least one concrete performance figure. The design choices are conventional in the best sense — ERC-721, IPFS, OpenZeppelin, MetaMask — which is what makes the work reproducible by others.
What remains unknown is the rest of the performance picture. The full latency results across file formats, the cost comparison the gas figure was paired with, and whether the framework has been deployed outside a testnet are not available in the published summary. The claimed advantages over existing platforms — better interoperability, lower energy use, stronger ownership control — are the authors' framing of the design, not independently measured outcomes. Readers weighing any token-related activity should also remember that crypto assets can lose most or all of their value quickly, and that smart contract code, however reviewed, carries execution risk the authors themselves list as a persistent industry problem.
