Dune Analytics: the underlying data can't lie, but nobody's checking whether the chart built on top of it does.
This is our second look at Dune Analytics, this time through the Data Aggregators lens: how it sources and structures data, rather than the screening-tool angle we used in our Data Screeners series. Real, disclosed genuine strengths: Dune indexes raw, immutable blockchain data into normalized databases via a purpose-built SQL engine, with infrastructure-grade products streaming that data directly into Snowflake, BigQuery, and Databricks. What we don't think should be set aside: the interpretation layer built on top of that verifiable data, every dashboard, every metric definition, is fully crowdsourced with no centralized review, and independent guidance warns plainly that an accurate blockchain record can still produce a misleading dashboard if the underlying definitions are sloppy. We weighted all of it below.
Our take, up front: Dune Analytics is a blockchain data platform founded in 2018 by Fredrik Haga and Mats Olsen, having raised roughly $79 million from Coatue, Union Square Ventures, and Dragonfly Capital, with no exchange investor found. This review uses a different scorecard than our earlier Data Screeners look at this platform: here we're weighing it specifically as a data aggregator, on sourcing methodology, coverage, and manipulation safeguards, rather than screening-tool depth. Real, disclosed genuinely rigorous underlying architecture: Dune "indexes and decodes blockchain data into relational databases" using DuneSQL, described as "the only query engine purpose built for crypto data analysis," built on the Trino engine, with over 100 blockchains indexed and more than 3 petabytes of data. Real, disclosed genuinely infrastructure-grade product line beyond dashboards: DataShare streams Dune's blockchain data directly into a customer's own Snowflake, BigQuery, or Databricks warehouse, alongside normalized vertical datasets covering stablecoin supplies, prediction markets (Polymarket, Kalshi), tokenized real-world assets, and perpetuals venues like Hyperliquid HyperCore. Real, disclosed genuine strength specific to this category: the underlying blockchain data itself is immutable and independently verifiable, meaning Dune cannot itself fabricate the raw record the way a self-reported-volume aggregator theoretically could. Real, disclosed structural limitation we want to present precisely: every dashboard and metric definition built on top of that raw data is fully crowdsourced, with "the accuracy of insights depends on the quality of the underlying queries," and independent guidance warns directly that "a wallet interacting three times in one day is not necessarily three users," that "a spike in volume may not mean adoption if one market maker generated it," and that a token holder count "can be misleading if dust wallets dominate," all real risks even when every underlying data point is completely accurate. Real, disclosed inconsistent self-reported blockchain-coverage figures across Dune's own current pages: 70-plus, 100-plus, and 130-plus, depending on which page and date. We weighted all of it below.
Real, disclosed genuinely rigorous underlying-data architecture: Dune indexes and decodes raw, immutable blockchain data into normalized relational databases via DuneSQL, a purpose-built query engine, with infrastructure-grade warehouse-delivery products (DataShare) reflecting genuine data-engineering depth. What tempers this significantly: the interpretation layer, every dashboard and metric definition, is fully crowdsourced with no centralized review, and independent guidance warns directly that accurate raw data can still produce misleading metrics without careful definitional work. Real, disclosed inconsistent self-reported blockchain-coverage figures across Dune's own current pages (70+, 100+, and 130+, depending on source and date).
Pros
- The underlying blockchain data is immutable and independently verifiable, not self-reported by any single party
- DataShare's direct warehouse integration reflects genuine, disclosed infrastructure-grade engineering
Cons
- No centralized review exists for the metric definitions built on top of the raw data
- Independent guidance directly warns that accurate data can still yield misleading dashboards
- Blockchain-coverage counts vary across Dune's own current pages
Real, disclosed massive, specific infrastructure scale: 100-plus blockchains indexed, 3-plus petabytes of data, and dedicated vertical datasets for stablecoins, prediction markets, tokenized real-world assets, and perpetuals (Hyperliquid HyperCore), plus ERC-20 coverage across 23 EVM chains.
Pros
- 3+ petabytes of indexed data reflects genuinely massive, disclosed technical scale
- Dedicated vertical datasets (RWAs, prediction markets, perpetuals) add real, distinctive breadth
Cons
- The exact blockchain count is one of the specific figures that varies across Dune's own materials
Real, disclosed genuine strength: the underlying blockchain data itself is immutable and independently verifiable, meaning the platform cannot itself fabricate the raw record the way a self-reported-volume aggregator theoretically could. What tempers this: metric definitions built on top of that raw data, active users, revenue, retention, are not standardized or centrally reviewed, and a single misleading dashboard can look authoritative without any single data point being factually wrong.
Pros
- No party, including Dune itself, can alter the underlying on-chain record being queried
Cons
- Ambiguous metric definitions (what counts as an "active user") can mislead without any data being inaccurate
Real, disclosed genuinely distinctive normalized dataset products functioning as de facto benchmarks (stablecoin supply and transfer tracking, prediction-market datasets), plus community dashboards frequently cited by major outlets, per Dune's own marketing, as "the real number."
Pros
- Normalized stablecoin and prediction-market datasets function as genuinely useful standing benchmarks
Cons
- Being "cited by major outlets" doesn't mean a given community dashboard was independently vetted first
Real, disclosed genuinely steep accessibility barrier: independent reviewers state directly that "without SQL knowledge Dune remains a viewing tool," with no mobile app, though free dashboard viewing itself is broadly accessible without an account.
Pros
- Thousands of public dashboards are viewable for free without requiring an account
Cons
- Building or meaningfully verifying a custom query requires real SQL fluency
Real, disclosed frequently-changing pricing, with independent reviewers noting directly that "plan tiers and pricing change several times a year," a real, disclosed forecasting difficulty for individual users and enterprise data-warehouse customers alike.
Pros
- A genuine free tier for dashboard viewing and limited query credits exists
Cons
- Frequent, disclosed pricing changes make budgeting for query credits or warehouse delivery genuinely harder
Real, disclosed genuinely large, active community-dashboard ecosystem, plus a free "Digital Asset Brief" research newsletter offered directly by Dune.
Pros
- A large, active dashboard community is a genuinely distinctive, disclosed extra found nowhere else in this series
Cons
- The same community-contribution model is the direct source of the disclosed interpretation-layer concern above
Trust the underlying data; verify the specific query before repeating a dashboard's headline number.
Given the real, disclosed integrity of Dune's underlying blockchain indexing, we'd treat any raw on-chain figure it surfaces as genuinely trustworthy, while opening the actual SQL behind any community dashboard's headline metric, or finding a second dashboard that defines the same term independently, before repeating that number as if it were an agreed-upon industry figure.
A second look, and the same core tension shows up in a sharper, more precise form.
Our earlier Data Screeners review of Dune focused on how maximally flexible its SQL access is for the people who can use it. Looking specifically at aggregation methodology this time, we found the more precise version of that same tradeoff: the actual blockchain data Dune indexes is about as trustworthy as data gets in this industry, immutable, public, and independently checkable by anyone with the SQL to look. That's a genuinely different foundation than a self-reported-volume aggregator, and it deserves real credit. But an aggregator's job doesn't end at storing the raw facts; it's also supposed to help you interpret them, and that's exactly the layer Dune leaves entirely to volunteers. A wallet transacting three times isn't necessarily three users. A volume spike from one market maker isn't adoption. These aren't edge cases, they're exactly the kind of everyday interpretive choices a widely-cited community dashboard makes every day, without anyone checking the choice was reasonable. We think that's a real, disclosed structural gap between "the data is honest" and "the dashboard is right," and it's worth scoring as its own thing rather than folding it into how impressive the underlying infrastructure is.
The scorecard above is deliberately general. Whether Dune Analytics is right for you depends heavily on which of these you already are.
The data engineer or analyst who wants direct, verifiable access to raw, indexed blockchain data at massive scale
This is exactly where Dune's real, disclosed underlying-data integrity delivers genuine, distinctive value.
The reader who opens a dashboard's actual SQL query before repeating its headline metric as an agreed-upon figure
Given the real, disclosed unreviewed interpretation layer we found, this specific habit genuinely matters here.
The enterprise team interested in DataShare for streaming blockchain data directly into an existing warehouse
This is a real, disclosed, genuinely useful infrastructure product worth evaluating on its own separate merits.
Anyone repeating a community dashboard's metric as an agreed-upon industry figure without checking how it defines its terms
A real, disclosed lack of centralized review over metric definitions makes this a poor fit for that specific assumption.
The scorecard covers the headline judgment calls. These four tables cover the specifics we didn't want to bury in prose; the infrastructure product line explained, the interpretation-layer risk quoted directly from independent guidance, how this score compares to our earlier Data Screeners review, and how all eight aggregators in this series compare so far.
The infrastructure product line, explained
| Product | What it does |
|---|---|
| Dune (core) | Web-based SQL editor and dashboard builder against indexed blockchain data |
| DataShare | Streams Dune's blockchain data directly into a customer's own Snowflake, BigQuery, or Databricks warehouse |
| Normalized vertical datasets | Purpose-built datasets for stablecoins, prediction markets, tokenized RWAs, and perpetuals venues |
| Dune API | Programmatic access to query results and blockchain data for building applications |
We're listing these separately because Dune's product line has genuinely expanded beyond dashboards into infrastructure-grade data delivery, a real, disclosed evolution worth understanding on its own terms.
The interpretation-layer risk, quoted directly from independent guidance
| The raw fact | Why it can still mislead |
|---|---|
| A wallet transacted three times in one day | "Is not necessarily three users" |
| Trading volume spiked | "May not mean adoption if one market maker generated it" |
| A token has many holders | "Can be misleading if dust wallets dominate" |
We're quoting these directly because they're the clearest, most concrete illustration we found of the actual mechanism behind our Data Sourcing score; every underlying fact here is completely accurate, and the metric built on top of it can still mislead.
How this compares to our earlier Data Screeners review
| Series | Lens | Score |
|---|---|---|
| Crypto Data Screeners | Maximally flexible SQL screening depth for technical users | 64.75/100 |
| Crypto Data Aggregators (this review) | Underlying data integrity vs. unreviewed interpretation layer | 63.5/100 |
We're showing both because the scores land close together for a reason: Dune's core tradeoff, powerful raw access paired with a crowdsourced, unreviewed interpretation layer, shows up clearly under either lens, just framed around a different primary question.
All eight aggregators, so far in this series
| DeFiLlama | CoinGecko | TradingView | CoinMarketCap | Dune Analytics | CoinPaprika | Messari | CoinCodex | |
|---|---|---|---|---|---|---|---|---|
| Core disclosed issue | Founder publicly defended Aave TVL figures against 2026 looped-liquidity inflation claims; separate unresolved governance dispute | Reported active exploration of a ~$500M sale, an open question over its founding independence claim | No default blended price; readers must select a composite symbol themselves | Exchange-ranking weights are undisclosed, with a documented history favoring owner Binance | Underlying data is verifiable, but dashboard metric definitions are fully crowdsourced with no centralized review | Self-contradictory scale and positioning claims across its own materials; a same-brand wallet app with serious complaints | Sells "amplification"-branded IR services directly to the same protocols it rates in "independent" research | Flagship AI price-prediction feature independently described as "opaque" and "more speculative than reliable" |
| Ownership | Launched 2021; no VC funding historically; May 2026 internal governance dispute over a proposed token | Founder-led since 2014; historically no VC funding; reported sale exploration as of Jan 2026 | Privately held, venture-backed since 2011, no exchange ownership found | Owned by Binance since April 2020 | Founder-led since 2018 (Fredrik Haga, Mats Olsen); ~$79M raised, no exchange investor found | Founder-led since 2018; small seed round (Manta Ray Ventures, Garvest Real Estate); no exchange investor found | Founded 2018; two CEO transitions since 2024; acquired by Blockworks June 2026 at a steep markdown | Founder-led since 2017; small team (~20 employees); no exchange investor found |
| Core strength | Independently-validated as the most methodology-transparent DeFi tracker in an academic survey of named competitors | A genuinely detailed, recently-strengthened anti-manipulation methodology (Trust Score Basilisk) | Genuinely transparent per-exchange labeling and a massive, actively-expanding feed network | A genuinely sophisticated, technically detailed anti-manipulation methodology | Immutable, independently-verifiable underlying data plus infrastructure-grade warehouse delivery (DataShare) | A specific, named proprietary outlier-filtering algorithm and a disclosed order-book Transparency tool | The deepest fundraising and investor database found anywhere in this project, plus a distinctive five-dimension risk framework | Genuinely broad, free, no-account-required access across crypto, stocks, forex, and metals |
| Our score | 74.5/100 | 70.25/100 | 65.5/100 | 64.25/100 | 63.5/100 | 58.5/100 | 57.5/100 | 55.5/100 |
Dune's score reflects genuine credit for arguably the most trustworthy underlying data source in this project, tempered by a real, disclosed structural gap between that data's integrity and the unreviewed interpretive layer built on top of it.
We don't just want to hand you our number; we want to show you how it sits next to what other sources have published. We're flagging directly that "best on-chain data tool" round-ups and developer sentiment tend to credit Dune's underlying data integrity and flexibility very heavily, while giving considerably less direct weight to the crowdsourced interpretation-layer risk than our methodology does for this specific category.
Our score lands modestly below the aggregated industry average; developer and technical sentiment credits Dune's genuine data integrity and flexibility very heavily, while giving considerably less direct weight to the unreviewed dashboard-interpretation risk we found central to this review.
| Source | Score | Type |
|---|
Scores compiled by our editorial team from publicly available reviews as of August 2026. "Editorial estimate" means the outlet didn't publish a single numeric score, so we converted their published verdict and sentiment into a comparable 100-point figure. Verify current figures directly with each source before citing them elsewhere.
Fredrik Haga and Mats Olsen founded Dune in 2018 in Oslo, Norway. The company has raised roughly $79 million from investors including Coatue Management, Union Square Ventures, and Dragonfly Capital, with no exchange-investor conflict found.
Dune indexes and decodes raw, public blockchain data directly, using a purpose-built query engine called DuneSQL, built on Trino, and normalizes it into relational databases spanning 100-plus blockchains and over 3 petabytes of data.
The underlying blockchain data itself is immutable and independently verifiable by anyone. The real, disclosed risk sits one layer up: the metric definitions and dashboards built on top of that data are fully crowdsourced and not centrally reviewed for accuracy or reasonableness.
An infrastructure product that streams Dune's indexed blockchain data directly into a customer's own data warehouse, such as Snowflake, BigQuery, or Databricks, for teams that want to build their own analysis on top of it.
Independent guidance recommends checking a dashboard's last update date, reviewing the author's track record, and comparing multiple dashboards on the same topic before trusting a single one's conclusions, since accuracy depends entirely on the quality of the underlying query.
To build your own custom queries, yes. Without SQL knowledge, Dune functions mainly as a free dashboard-viewing tool; thousands of existing community dashboards are viewable without an account.
Dashboard viewing is free. Running custom queries uses limited free-tier credits, with paid tiers for higher allowances, API access, and DataShare warehouse delivery. Pricing reportedly changes several times a year, so check Dune's own pricing page for current figures.
Our Data Screeners review (64.75/100) focused on Dune's flexibility as a screening and query tool. This Data Aggregators review (63.5/100) focuses specifically on data sourcing and integrity, distinguishing between the platform's trustworthy underlying data and its unreviewed crowdsourced interpretation layer.
More Reviews
Kaiko – Crypto Data Aggregator Review
Score: 60.0/100. Massive institutional coverage and a real frontrunning catch, but its new investors include an exchange and two trading firms.
Read More
CoinMarketCap – Crypto Data Aggregator Review
Score: 64.25/100. A sophisticated anti-manipulation methodology undercut by undisclosed ranking weights and a history favoring owner Binance.
Read More
Glassnode – Crypto Data Aggregator Review
Score: 72.0/100. Sophisticated entity-adjusted, point-in-time on-chain methodology, published transparently, with a rare clean, founder-led ownership picture.
Read More
TradingView – Crypto Data Aggregator Review
Score: 65.5/100. Transparent per-exchange labeling and a massive feed network, but no default blended price without user selection.
Read More