Solana Datasets for Academic & Quant Research: Citation-Ready
On-chain datasets built for reproducible research: SHA-256 checksums, fixed temporalCoverage windows, identity-scrubbed pseudonymous rows, and schema.org/Dataset JSON-LD that makes every set citable and discoverable in Google Dataset Search.
MadeOnSol·· 7 min read
Disclosure: This article contains affiliate links. If you sign up through them, MadeOnSol may earn a commission at no extra cost to you. This never affects our rankings, ratings, or reviews.
Building a product on Solana data?
Skip the Geyser pipeline — embed KOL flow, deployer reputation, and the all-DEX firehose over REST, WebSocket, or webhooks. The Business tier (€400/mo) is the self-serve embed license at 500k calls/day — Enterprise adds white-label & redistribution.
Before you commit, pull the free 5,000-row sample for any set. It carries the real schema and a representative slice of the distribution, so you can validate column types, check for the variables your model needs, and sanity-check ranges without spending anything. Reviewers appreciate that you inspected the data before building on it.
In your methods, cite the dataset by its SKU, its declared coverage window, and the SHA-256 hash of the exact file you analyzed. That triple — name, window, hash — is enough for full reproduction. Because the pages carry schema.org/Dataset metadata, you can also point a machine-readable citation straight at the source URL.
Pick the population deliberately. The four sets answer different questions, and matching your research question to the right coverage window is the first methodological decision:
The at-execution pricing in sol-kol-trades deserves a specific note for quants: each row stamps price_usd_at_trade and market_cap_usd_at_trade at the moment of the trade, so you can compute realized entry and exit without reconstructing a price oracle. That removes a whole class of look-ahead and interpolation error from event studies on labelled trades.
FAQ
Are these datasets reproducible for academic work?
Yes. Every set is an immutable CSV.gz with a SHA-256 checksum and a fixed, declared temporalCoverage window. You cite the file by name, window, and hash; anyone with the same file can verify it byte-for-byte and re-run your analysis. That is reproducible by construction, unlike a live API or a "latest" export.
Is it ethical to use on-chain wallet data in research?
The data is derived from a publicledger and is identity-scrubbed and wallet-keyed — pseudonymous public addresses with their on-chain metrics, no names, personas, or off-chain identifiers. That keeps it in pseudonymous-public-data territory suitable for research, and lets you publish aggregate results without exposing an individual.
Tools mentioned
Compare features and read reviews.
Live health scores, average ratings, and direct links on MadeOnSol.
MadeOnSol isn't an RPC or another generic token API — it's Solana memecoin intelligence that's painful to build in-house, pulled from dual-region gRPC shred streams: sub-second from the on-chain event to your app.
Signals you'd otherwise build
KOL & smart-money flow from 1,000+ labeled wallets, deployer reputation, coordination clusters, linked-wallet entity resolution, and an all-DEX firehose.
Embed it in your product
The Business tier (€400/mo, self-serve) licenses you to display MadeOnSol data inside your own product — 500k calls/day, 10 WS + 5 firehose connections. Enterprise above adds white-label & redistribution rights and custom endpoints.
Evaluate first, commit later
Test everything on a free ULTRA key — no commitment. If it fits, we scope volume or white-label pricing.
For academic and quant work, an on-chain dataset is only usable if it is reproducible, ethical, and citable. Ours are built for all three: every set ships as CSV.gz with a SHA-256 checksum so you can prove the file you analyzed is the file you cite, a fixed temporalCoverage window so the sample never silently shifts under you, identity-scrubbed pseudonymous rows (wallet-keyed, no names or personas) so you can publish without a personal-data problem, and each dataset page emits schema.org/Dataset JSON-LD — making the set eligible for Google Dataset Search and machine-readable citation. This post is the research-integrity guide across the whole catalog: what makes these datasets defensible in a paper, a thesis, or a reproducible backtest.
If you want a specific set, the deep-dives are linked below. If you want the buy page, start at the datasets hub or take everything in the full-access bundle (€499, all four sets). What follows is the methodology contract, not a product tour.
The four properties that make a dataset citable
Property
What we ship
Why it matters for research
Integrity
SHA-256 checksum per file
Prove file identity; detect corruption or tampering
Fixed window
Declared temporalCoverage
The sample doesn't drift; results are reproducible
Ethics
Wallet-keyed, identity-scrubbed
Pseudonymous data; publishable without PII exposure
Discoverability
schema.org/Dataset JSON-LD
Google Dataset Search + machine-readable citation
Evaluation
Free 5,000-row sample
Inspect schema and distribution before committing
Reproducibility: the checksum is the point
The failure mode of on-chain research is that the underlying data moves. A live API returns different rows tomorrow; a "latest" export is a different file next week. Neither is reproducible, and a reviewer cannot re-run your analysis against a moving target. A checksummed, immutable CSV fixes this: you record the SHA-256 hash in your methods section, and anyone with the same file can verify byte-for-byte that they are working from your exact input. This is the difference between "we queried an API" — unrepeatable — and "we analyzed file X, hash Y" — repeatable by construction.
Fixed temporalCoverage: no silent drift
Each dataset declares an explicit coverage window, and that window is part of the published schema.org/Dataset metadata:
sol-smart-money (€149): ~1.25M scored wallets, coverage from 2026-04-12. See the for the ML angle.
sol-kol-trades (€149): 1.5M+ KOL trades, coverage 2026-05-15 → present — the fully-priced era, where each row carries the price and market cap stamped at execution.
sol-deployers (€99): ~27k real Pump.fun deployers; a current-state reputation snapshot whose earliest launches reach back to January 2024.
Because the window is declared and the file is immutable, your sample definition is unambiguous. A reader knows exactly what population your results generalize to, and there is no risk that a later refresh quietly changed the denominator.
Ethics: pseudonymous by construction
Every row is wallet-keyed and identity-scrubbed — public on-chain addresses and their derived metrics, with no names, personas, or off-chain identity attached. This is a deliberate design choice, not an afterthought. On-chain addresses are pseudonymous public data; by never joining them to real-world identity, the datasets stay on the right side of the line for research use, and you can publish aggregate results without exposing an individual. For an IRB or an ethics review, the relevant statement is simple: the data is derived from a public ledger, keyed on pseudonymous addresses, and contains no personal identifiers.
Discoverability: schema.org/Dataset JSON-LD
Each /datasets/<sku> page emits structured schema.org/Dataset markup including temporalCoverage and variableMeasured. Practically, that means the sets are eligible for Google Dataset Search and are machine-readable to the tools researchers and AI systems use to locate and cite data. A dataset that describes itself in a standard vocabulary is one a citation manager, a search index, or a language model can reference precisely — the same discipline that gets a dataset cited rather than merely downloaded.
Can these datasets be found in Google Dataset Search?
Each /datasets/<sku> page emits schema.org/Dataset JSON-LD with temporalCoverage and variableMeasured, which makes the sets eligible for Google Dataset Search and machine-readable to citation tools and AI systems. The structured metadata is what makes them discoverable and citable rather than just downloadable.
Which dataset should a quant start with?
Match the question to the window. For wallet-skill or bot-classification ML, use sol-smart-money. For realized entry/exit event studies with no oracle needed, use sol-kol-trades. For deployer-outcome work, use sol-deployers. For a full L2 from block zero, use rhc-intelligence. Grab the free 5,000-row sample of any of them first.
Bottom line
For research, the specification matters more than the row count. These datasets are checksummed (reproducible), fixed-window (no drift), identity-scrubbed (publishable), and schema.org-described (citable and discoverable). Cite them by SKU, coverage window, and SHA-256 hash, and validate against the free 5,000-row sample first. Browse all four at /datasets, or take the whole citable corpus — Solana smart money, KOL trades, deployers, and Robinhood Chain from genesis — in the full-access bundle. For the catalog overview, start at the datasets hub.
