On 12 June 2026 an OpenAlex pull tagged 128,286 papers as blockchain or cryptocurrency between 2013 and mid-2026. Yearly output in that file went from 176 papers in 2013 to 20,668 in 2025. That is 117 times more papers in twelve years. The tables are in Web3 Research Trends on SSRN.
A paper total is not a count of software that still runs. Working chains, court filings, and statutes are a separate record.
Three counts, not one field
A keyword jump can look like a boom. A tracker row can look like a currency that already exists. The June file is one tagged set of papers. It is not value locked, and it is not a list of currencies that exist.
The tagged set is 128,286 papers with those two OpenAlex tags, pulled on 12 June 2026. A wider abstract search in the same paper finds 1,938,409 papers. That is an upper bound, not the tagged set. Value locked, court dockets, statutes, CBDC trackers, and shipping platforms sit on a different meter. Abstract counts can rise while those stay put.
Priem, Piwowar, and Orr (2022) built OpenAlex as an open replacement for Microsoft Academic Graph.
The yearly bars in the June chart add up to 127,286. The tagged total is 128,286. The paper does not explain the 1,000-record difference.
| Query | Date | Count | What it is |
|---|---|---|---|
| Tags C2779687700 or C180706569, 2013 to mid-2026 | 12 Jun 2026 | 128,286 | Tagged OpenAlex set in SSRN 6990919 |
| Yearly bars in that chart, summed | 12 Jun 2026 | 127,286 | Same snapshot, 1,000 short of the tagged total |
| Abstract keyword search for Web3 words | 12 Jun 2026 | 1,938,409 | Wider net, not the tagged set |
| Crossref | 12 Jun 2026 | 124,333 | Coverage check, within 3% of the tagged set |
The June paper also reported other databases. Europe PMC 61,191. arXiv 11,312. DBLP 37,246. Journal articles are 68.9% of the tagged set, book chapters 13.9%, preprints 8.2%. Open access is 50.2% here. Piwowar, Priem, Larivière, et al. (2018) put all-field Crossref journal articles at 28% open access. That 28% is 2018, all fields.
Earlier counts were smaller because the field was smaller. Miau and Yang (2018) stopped in 2017. Dabbagh, Sookhak, and Safa (2019) swept IEEE Access. Casino, Dasaklis, and Patsakis (2019) classified applications. Ante (2020) counted smart-contract papers. The June file is about 100 times larger than those studies. Size still measures papers, not deployments.
What people say, and what the record is
| Public claim | Source | What the record actually is |
|---|---|---|
| 128,286 blockchain papers | SSRN 6990919, 12 Jun 2026 | Tagged papers that day. Yearly bars sum to 127,286. |
| DeFi mentions grew 74x | June abstracts, 2019 to 2025 | The word in an abstract. Not dollars locked. |
| 11,418 supply-chain papers | June bigrams | Abstracts. TradeLens shut in 2022. |
| 205 CBDC papers in 2025 | June file | Papers. Not 205 live currencies. |
| 146 countries on a CBDC | Atlantic Council, May 2026 | In the tracker, including research. 3 retail launches. e-CNY is still a pilot. |
| $35T stablecoin volume | CoinDesk / McKinsey / Artemis, 2025 | On-chain transfers. Genuine payments printed as both $390B and $380B. |
| China leads the literature | June geo, 17,258 papers | Author country. PBOC banned virtual-currency business in 2021. |
| India is close second | June geo, 17,037 papers | Author country. 30% tax on VDA transfer. Chainalysis 2023 everyday-use #1. |
What the papers were about, year by year
2013 to 2016. 176 to 955 papers. Most of the work was still Nakamoto (2008): stop double-spending without a bank. Lamport, Shostak, and Pease (1982) asked how generals agree when some messengers lie. Those four years in the chart sum to 2,111 papers, about five weeks of 2025 output.
2017 to 2019. 2017 grew 164% to 2,520. 2018 grew 178% to 6,995, the fastest year in the June file. Buterin's Ethereum whitepaper and Wood's Yellow Paper turned a ledger into a computer that stores state. Androulaki et al. (2018) described Hyperledger Fabric for firms that wanted a chain only invited members can write to. By 2019 the yearly count was 10,591.
2020 to now. Growth slowed. 2024 grew 1.8%. Papers from 2020 through mid-2026 are 82.7% of the June file (106,069 of 128,286). The 2026 slice in that share is 8,419 papers seen by 12 June, not a full year. Ethereum's Merge on 15 September 2022 ended proof-of-work. The original Foundation post for that day is gone. The live record is ethereum.org/roadmap/merge. ethereum.org's energy page cites CCRI: about 2,601 MWh a year after the switch (0.0026 TWh), a drop of more than 99.988% in electricity and about 99.992% in carbon against proof-of-work. Those are estimates, not an official Foundation promise. The comparison table on that page, with figures dated July 2023, puts Bitcoin at 149 TWh/yr, about 53,000 times proof-of-stake Ethereum. Bitcoin demand moves daily. The live index is the Cambridge Bitcoin Electricity Consumption Index.
Supply-chain papers kept rising after TradeLens shut down
"Blockchain" is in 68.5% of June abstracts (87,818). "Security" is in 46.3% (59,388). Zheng, Xie, Dai, Chen, and Wang (2018) listed failure modes early. Khan and Salah (2018) did the same for IoT. Security stayed in the abstracts. "Smart contract" is in 29.9% (38,414). "Bitcoin" is in 27.9%.
"Blockchain technology" (44,102) and "blockchain based" (18,516) are how authors add a ledger to some other system. "Supply chain" is the largest non-finance job at 11,418. Saberi, Kouhizadeh, Sarkis, and Shen (2019) tied it to sustainability claims. Christidis and Devetsikiotis (2016) tied it to IoT.
The largest industry test of that claim was TradeLens, a Maersk and IBM shipping platform announced in 2018. On 29 November 2022, Maersk said it would shut the platform. The 19 December 2022 Wayback capture still has the text. The companies wrote that they had built a working platform, but "the need for full global industry collaboration has not been achieved," and it had not reached "the level of commercial viability necessary to continue work and meet the financial expectations as an independent business." The plan was to go offline by the end of Q1 2023. Supply-chain papers kept arriving after that. The tagged set still includes papers about products that later shut.
Health and privacy have early papers and smaller counts. Zyskind, Nathan, and Pentland (2015) stored pointers to personal data, not the data. Ekblaw, Azaria, Halamka, and Lippman built MedRec. "Healthcare" is in 13.7% of abstracts. "Machine learning" (5,667) plus "artificial intelligence" (5,463) is 11,130 mentions. Kosba, Miller, Shi, Wen, and Papamanthou (2016) hid contract internals (Hawk). Lu, Huang, Zhang, Maharjan, and Zhang (2020) put federated learning on a ledger for vehicles. Mentions in 11,130 abstracts are not 11,130 working systems.
DeFi papers rose while locked dollars split
DeFi mentions: 7 papers in 2019, 12 in 2020, 63 in 2021, 516 in 2025. That is 74 times in six years. NFT mentions: 60 in 2021, 426 in 2023, 334 in 2024, 366 in 2025.
Uniswap's docs still describe the pool that replaced the order book. Live value locked is a different meter. DefiLlama is the public index. Its 2025 sector note does not give one number for "DeFi." It splits:
- Lending TVL rose from $48.15 billion to $64.06 billion inside 2025. Top ten protocols held $57.00 billion, 89.0% of lending TVL.
- Liquid staking slipped from $58.73 billion to $57.62 billion.
- Restaking fell from $23.85 billion to $18.78 billion, a 21.2% drop.
- Tokenised US Treasuries and money-market products rose from $3.87 billion at end-2024 to $9.12 billion by late 2025.
A paper count can rise while some locked-dollar series rise and others fall.
The algorithmic stablecoin that blew up is in a court file. It is also in abstracts. On 16 February 2023 the SEC charged Terraform Labs and Do Kwon with a multi-billion-dollar securities fraud around UST, LUNA, and related tokens. Enforcement Director Gurbir S. Grewal said, as alleged, "the Terraform ecosystem was neither decentralized, nor finance." On 5 April 2024 a jury found them liable for securities fraud. On 13 June 2024 the SEC announced a package above $4.5 billion: Terraform $3,586,875,883 disgorgement, $466,952,423 interest, $420 million penalty; Kwon $110 million disgorgement and $14,320,196 interest, joint and several with Terraform, plus an $80 million penalty. That release says the May 2022 de-peg "wiped out $40 billion in market value nearly overnight." Academics kept writing about DeFi after that. They wrote about reserved dollars and over-collateralized loans, not about UST.
NFT paper volume peaked in 2023, after the speculative market had already broken. Later papers are more about who owned an item than about profile pictures. The word stayed in abstracts after floor prices died.
Stablecoins already move money. The payment claim is smaller than the chain volume. CoinDesk, citing McKinsey and Artemis (23 January 2026), put 2025 on-chain stablecoin volume above $35 trillion. The same article prints $390 billion of "genuine" payments in the summary and $380 billion in the body ($226 billion B2B, $90 billion payroll and remittances, $8 billion capital markets). Those two figures do not match. Genuine payments are about 1% of the $35 trillion either way. McKinsey put the global payments market above $2 quadrillion, so even the higher figure is about 0.02% of that total. Circle and Tether still dominate issuance.
Keywords that rose, and what actually shipped
Growth ratios in the June file compare 2025-2026 mentions to 2022-2023. A 6x jump off 37 papers is not the same as a 3x jump off 4,000.
Post-quantum. "Quantum resistant" grew 6.6x (37 to 243). "Post-quantum" grew 3.9x (95 to 367). Together, 610 papers in 2025-2026. Bitcoin and Ethereum still sign with elliptic curves. Shor (1994) showed a large quantum computer would break those signatures. There is no such machine on the public record. What moved the papers is a standards event. On 13 August 2024, NIST published FIPS 203, 204, and 205. FIPS 203 is ML-KEM. The CSRC PQC page tells operators to start migrating. It does not give a date when ECDSA dies.
Scale. "Layer 2" grew 4.2x (275 papers). "zk rollup" grew 3.4x off 34 papers. The count rose, but 3.4x off 34 papers is not a volume story. Buterin's 2021 rollup note still describes the method: execute off the base chain, post data or a proof back. L2BEAT tracks live rollups. "Zero knowledge" grew 2.1x (743), the fastest rise among cryptographic primitives in the June file.
Rules and tokens. "Regulatory" grew 3.5x and the count is large: 4,529 papers. "Real world asset" grew 5.6x off 89 papers. "Stablecoin" grew 2.7x (576). "DeFi" grew 2.6x (808). "NFT" and "metaverse" sit at 0.8x.
BlackRock launched BUIDL on Ethereum on 20 March 2024, a tokenized money-market fund for qualified investors, $5 million minimum, cash and Treasuries and repo, tokens targeting $1. Eighty-nine "real world asset" papers are not 89 copies of that fund. DefiLlama's tokenised-Treasury line ($3.87 billion to $9.12 billion in 2025) is closer to live dollars.
EU and US rules arrived after the paper wave
The EU's Markets in Crypto-Assets Regulation (Regulation (EU) 2023/1114) was adopted on 31 May 2023 and published on 9 June 2023. It is the first EU-wide set of rules for issuance and crypto-asset services.
The US followed on stablecoins. The GENIUS Act became Public Law 119-27 on 18 July 2025. Payment stablecoins, 1:1 reserves. Signing remarks are in the Daily Compilation. The statute covers reserved dollars. It does not cover the algorithmic design the SEC sued.
Spot Bitcoin exchange-traded products were approved on 10 January 2024. Chair Gensler's statement is the public-market event. He wrote that the Commission did not approve bitcoin itself.
146 countries are on a CBDC tracker. Three have launched.
The Atlantic Council tracker (May 2026) has 146 countries and currency unions with a CBDC entry, covering over 98% of world GDP, up from 87 in May 2022. 77 are in development, pilot, or launch. 41 pilots. Three retail CBDCs have launched: the Bahamas, Jamaica, and Nigeria. Every G20 member except the US is in the tracker. Canada, Australia, and Norway have stepped back from retail. The BIS Innovation Hub CBDC page covers the technical work. The ECB digital euro page is the euro-area project. India's RBI concept note set out an e-rupee. CBDC papers in the June file go from 1 in 2016 to 205 in 2025. That follows official work. It is not 205 live currencies.
The tracker splits "launch" from "pilot." High volume and a launch label are not the same status.
| Tracker bucket (May 2026) | Count | What it is |
|---|---|---|
| In the tracker for a CBDC | 146 | Research through launch. 98% of world GDP. |
| Development, pilot, or launch | 77 | "Advanced" in the tracker. |
| Pilots | 41 | Includes e-CNY, the largest by volume. |
| Retail CBDCs fully launched | 3 | Bahamas, Jamaica, Nigeria. |
By December 2025, according to the same tracker, retail e-CNY had processed more than 3.4 billion transactions worth about 16.7 trillion yuan (~$2.3 trillion). In January 2026 the PBOC reclassified e-CNY as deposit liabilities. The Bahamas, Jamaica, and Nigeria are the three full launches. e-CNY has the most volume and still sits in the pilot column.
Nigeria launched eNaira in October 2021. The IMF's May 2023 working paper (WP/23/104) found about 14,000 transactions a week, 98.5% of wallets unused in a given week, and average weekly value of 923 million naira, 0.0018% of M3. The Fund called public adoption "disappointingly low." A later Article IV staff report, as quoted in BusinessDay, put cumulative transactions at 854,512 and N29.3 billion against N611 trillion of electronic payments in Nigeria in 2023.
China is the opposite case. Cryptocurrency trading is illegal. The PBOC and nine other departments issued 银发〔2021〕237号 in September 2021: virtual currency has no legal-tender status, related businesses are illegal financial activity, overseas exchanges serving mainland residents are also illegal. The NDRC mining notice (发改运行〔2021〕1283号, 3 September 2021) ordered existing mines out and cut finance and power. The e-CNY is the state substitute. Atlantic Council, 15 January 2026, citing PBOC year-end data, matches the tracker on the 3.4 billion transactions and 16.7 trillion yuan. Project mBridge, after the BIS stepped back in October 2024, had processed about $55.49 billion, with e-CNY about 95% of settlement. That is a central bank payment system, not bitcoin.
The Blockchain-based Service Network introductory whitepaper (5 February 2020) describes a network for government and firm chains that invited members write to. Chinese paper volume can be high while "Bitcoin" as a keyword is no longer the center. Inside China the legal topic is not a public coin.
China and India write most of the papers. US products list in New York.
China 17,258 papers (13.5%). India 17,037 (13.3%). United States 12,488 (9.7%). Top three: 36.5% of the June file. In most computer-science fields the US is first or second. Here it is third.
OpenAlex assigns a paper to every co-author country, so shares can overlap. A paper with authors in China and the US counts in both.
The country split matches use more than it matches famous US universities. Chainalysis's 2023 Global Crypto Adoption Index put India first in everyday use. Nigeria second, Vietnam third, the United States fourth. The 2023 geography report put India second in raw estimated volume after the US, about $250 billion received in the year to June 2023. That was after section 115BBH of the Income-tax Act (Finance Act 2022): 30% on transfer of a virtual digital asset, no expense other than cost of acquisition, no loss set-off. Section 194S deducts 1% TDS. People used crypto anyway. Indonesia is rank 6 in the June paper file (3,106) and rank 7 on that adoption index. Saudi Arabia is rank 10 here (2,423).
No single US university sits in the top eight institutions in the June file. The Chinese Academy of Sciences has 924 papers. Indian Institutes of Technology occupy two of the top six slots. US work is spread. Paper volume is not a quality ranking. Citation-weighted lists can rank the same countries differently. The State of AI essay makes the same point on a five-million-paper pull.
China leads the paper count and bans the public coin. India is a close second in papers, first in everyday use, and taxes that use at 30%. The US is third in papers and first in the products that list on NYSE and Nasdaq.
Most of these papers will not be cited
43.5% of the 128,286 papers have zero citations. 31.5% have 1 to 13. 20% have 14 to 41. 5% have 42 or more. The 99th percentile sits at 249. Nakamoto (2008) is an outlier at 14,286 OpenAlex citations in the June pull. Zheng et al. (2017) is the survey a thousand later papers cite instead of citing the primary system. Melanie Swan's 2015 book still collects citations. Most later "blockchain based X for Y" papers collect none.
82.7% of the tagged set is from 2020 on. New papers have had little time. Average citations per paper in the yearly file fall from 85.0 in 2013 to 0.8 in 2025. Most of this file will never be read twice.
What the file counts
TradeLens shut after Maersk and IBM said it was not commercially viable. UST de-pegged in May 2022; the SEC later published a package above $4.5 billion. eNaira launched in 2021; the IMF found about 14,000 transactions a week and 98.5% of wallets unused. The June file still includes the papers written about each of them. It is 128,286 tagged works. 43.5% have zero citations. It is a literature pull, not a count of software that still runs.
China is 13.5% of that file and India is 13.3%. MiCA is the EU statute. The GENIUS Act is the US payment-stablecoin statute. The same literature is searchable in the Web3 Reports Library.
The 12 June 2026 tables are in Web3 Research Trends (SSRN 6990919).