The complete list of all websites, ranked.
The full list of all registered websites on the public web, 308,259,087 of them, each scored by PageRank and Harmonic Centrality. Country by IP and registration dates sit on every row, and you skip the trillion-edge link graph because the ranks arrive pre-computed.
✓ Includes All active domains columns · ✓ Daily fresh export · ✓ Free 10K sample
Inside the Websites + Ranking dataset: a live data sample
Five real rows from the live database, the exact 9-column schema you download, nothing mocked up.
| Domain | TLD suffix | TLD | Domain type | Country by IP | PR value | Harmonic value | Domain creation date | Domain expiration date |
|---|---|---|---|---|---|---|---|---|
| instagram.com | com | com | ICANN | DE | 0.0062957 | 26,877,164 | 2004-06-04 | 2034-06-04 |
| cloudflare.com | com | com | ICANN | US | 0.00590658 | 22,443,220 | 2009-02-17 | 2033-02-17 |
| youtube.com | com | com | ICANN | US | 0.00483818 | 25,770,158 | 2005-02-15 | 2027-02-15 |
| linkedin.com | com | com | ICANN | US | 0.00380516 | 24,467,486 | 2002-11-02 | 2026-11-02 |
| wordpress.org | org | org | ICANN | US | 0.00191729 | 21,394,334 | 2003-03-28 | 2035-03-28 |
Live values straight from the dataset, not placeholders. ↓ Free 10K preview
Two importance signals on every website
Every row in the list of all websites carries two independent measures of how much the rest of the web points at it. Use one, or use both to cross-check.
PageRank
Link-graph authority. A site that many important pages point at scores higher, so a one-line filter keeps only names above the threshold you set. Example, instagram.com sits near the top at a PR of 0.0063.
Harmonic Centrality
A second centrality measure of how central a site is to the whole graph. It is more robust to link spam than PageRank alone, which is why shipping both lets you rank and sanity-check in one pass.
Reading the two ranks
Both values ship raw, so you set the cutoff that fits your use. Here is roughly how each one scales, and how to turn it into a filter.
The PageRank scale
PageRank runs from tiny fractions upward. The busiest sites on the web sit near 0.006, a solid mid-tier site lands a couple of orders of magnitude below that, and the long tail trails toward zero. Pick a floor that matches the quality bar you want, and the list filters itself down to the sites worth your attention.
The Harmonic scale
Harmonic Centrality is a large integer that climbs with how many short paths across the graph lead to a site. The top of the web reaches the tens of millions, while ordinary sites sit far lower. It moves more smoothly than PageRank, so it is the better tie-breaker when a batch of sites bunch up at a similar PageRank.
Nine columns on every ranked website
The four-column domain spine from All active domains, plus five ranking and lifecycle fields.
How every website gets ranked, daily
The heavy graph compute runs on our side, so the ranks land in your copy already computed.
Crawl
Sweep the web for active sites.
Verify
Confirm each is live via DNS, HTTP, WHOIS and RDAP.
Link graph
Rebuild the web-scale link graph.
Score
Compute PageRank and Harmonic per site.
Export
Add country and dates, ship one JSONL file.
Who works from the list of all websites
One daily-refreshed, ranked file, many teams. Here is how different buyers turn it into an advantage.
Track ranking shifts over time
Daily exports are diffable, so you watch how PageRank and Harmonic Centrality move across the whole web, not just the sites a tool happens to sample.
Filter your market by authority
Pull the active websites in a country above a minimum PageRank and you have an instant addressable-market list with the junk already filtered out.
Score every link target in bulk
Skip per-lookup API limits and score every prospect in one pass over the file, ranked by the value a link from them would carry.
Weight inventory by real authority
Join bid requests against the list of all registered websites to weight or exclude domains by PageRank and by hosting country.
A ranked census of the web
Study the shape and authority distribution of the whole web on a dataset that refreshes every day, not a year-old crawl.
Spot undervalued names
Cross-reference PageRank with expiration dates to find high-authority sites near renewal across 239 countries.
Rank the web yourself, or buy it pre-computed
Building it yourself
- A web-scale link graph means storing and processing trillions of edges.
- PageRank over that graph is a heavy distributed compute job to run and re-run.
- Per-lookup SEO APIs cap you at a few thousand queries and meter every call.
- You still have to geolocate every IP and pull registration dates separately.
One file from webatla
- PageRank and Harmonic Centrality arrive pre-computed on every row.
- 308.3M ranked websites across 239 countries in a single file.
- Re-exported daily, so ranks and lifecycle dates stay current.
- No per-lookup metering, no rate limits.
Drop the ranked list into your stack
JSON Lines and zstandard, one object per row. It loads into Postgres, ClickHouse, BigQuery or DuckDB with no preprocessing.
SELECT domain, pr FROM 'websites-ranking.jsonl.zst'
WHERE country = 'DE' ORDER BY pr DESC LIMIT 100
One-line recipes
Because every rank and date sits in a plain column, the queries you actually want are one line each. A few of the lists teams pull most often.
Top sites in a market
A ranked list of the strongest sites in any country.
Outreach prospects
Every .io site worth an email, already scored for you.
High authority near expiry
A watchlist of valuable domains coming up for renewal.
Where the ranked web lives
A one-query breakdown of the web by hosting country.
Fresh names with early traction
A list of young sites that are already picking up links.
Compare two TLDs by authority
Average PageRank per TLD, side by side, to see where the strong sites cluster.
Ranks you can build on
Computed on our own graph
PageRank and Harmonic Centrality come from webatla's own web-scale crawl, not resold from a third party.
Live sites only
Every website is verified via DNS, HTTP, WHOIS and RDAP before it earns a row, so you rank real properties.
Rebuilt daily
The whole list is regenerated each day, so ranks and expiration dates never drift into a stale cache.
Spanning 1,435 TLDs and 239 countries
webatla tracks domains across every TLD and country. The largest by domain count, each linking to its full profile.
Subscription, freshness, licensing & refunds
How big is the list of all websites?
The current export holds 308,259,087 ranked, active websites across 239 countries and 1,435 TLDs. It refreshes every day.
Where do the ranks come from?
We compute PageRank and Harmonic Centrality over our own web-scale link graph, rebuilt with the daily export. Both raw values ship in every row.
What is the difference between PageRank and Harmonic Centrality?
PageRank measures authority flowing through links, so a site many important pages point at scores higher. Harmonic Centrality measures how central a site is to the whole graph and is more robust to link spam. Having both lets you rank and cross-check.
Can I filter the list by country or minimum rank?
Yes. Every row carries country by IP and both rank values, so a one-line WHERE clause keeps only sites above a PageRank threshold in the countries you care about.
How is this different from All active domains?
All active domains is the four-column list of every registered domain. Websites and Ranking keeps the active websites among them and adds country by IP, PageRank, Harmonic Centrality and registration dates.
How often do the ranks change?
The link graph is rebuilt with every daily export, so PageRank and Harmonic Centrality can move from one day to the next. Diff two snapshots to measure the shift for any site or any slice of the list.
Can I track ranking history over time?
Yes, by keeping the daily exports. Each file is a dated snapshot, so stacking them gives you a time series of PageRank and Harmonic Centrality for every website.
Do you cover ccTLD and non-English sites?
Yes. The list spans every TLD and ccTLD, and country is resolved by IP, so you can slice the ranked web by the market a site is served from rather than by language.
Which names count as a website here?
A registrable domain that resolves and serves a live site. Names that only answer DNS without a working site are not ranked, and live in All active domains instead.
Is the ranking global or per-country?
The ranks are global, computed over one worldwide link graph, so a PageRank in Japan is comparable to one in Brazil. Country by IP is a separate column, so you can still narrow the global list down to any single market.
Can I match the list against my own domains?
Yes. Every row is keyed on the registrable domain, so a hash join against your CRM or a list of competitors attaches PageRank, Harmonic Centrality, country and dates to each name in one pass.
What format does it come in?
JSON Lines, zstandard-compressed (6.1 GB), one documented schema. It loads into Postgres, ClickHouse, BigQuery or DuckDB with no preprocessing.
How does access work?
You get one month of access, or a year at €990. Inside that window you can re-download the latest daily export whenever you want.
How do I get the data after I pay?
The moment your payment clears, the dataset appears in your account downloads. You can re-download the latest daily export as often as you like for your whole one-month access window.
Can I use the data commercially?
Yes. Commercial use inside your organisation is covered by the webatla data licence (see the Terms). For redistribution, reselling or multi-year licensing, contact us first.
Do you offer refunds?
Datasets are digital products, so all sales are final, which is exactly why every dataset has a free 10K-row preview to check the schema before you buy. The one exception is a material defect such as a broken file, the wrong schema, or more than 5% missing rows. Email support within 72 hours and we'll re-issue or refund.
Get the ranked list of all websites.
308.3M ranked sites with PageRank and Harmonic · daily fresh export · €99 for a month.