Skip to main content
  • Pricing
Tutorials

Levels of the internet explained: surface web, deep web and dark web

By Oakgen TeamPublished 3 min read

Share
Levels of the internet explained: surface web, deep web and dark web

"Levels of the internet" usually refers to the layers of the web, and the standard version has three: the surface web, the deep web, and the dark web inside it. The Congressional Research Service (CRS), which prepares nonpartisan reports for Congress, defines the layers by whether content is indexed by search engines and whether special software is needed to reach it. Viral videos and lists add more numbered levels, but none of the sources opened for this guide define them, so this page sticks to what is documented.

Internet versus web

CRS points out that people treat the internet and the web as synonyms, but the web is only one portion of the internet. Email, file transfers and direct messaging also run over it. The layers below are layers of the web.

The three layers

LayerCRS definitionTypical example
Surface webContent that has been indexed and is accessible through traditional search engines such as GoogleA public news article or company site
Deep webContent that has not been indexed and so is not reachable by a traditional search engine; often needs authentication and permissionPrivate intranets at corporations, agencies or universities; commercial databases like Westlaw
Dark webThe segment of the deep web that has been intentionally hidden; generally cannot be accessed without special softwareSites reachable through Tor or I2P

Notice that "deep" mostly describes ordinary material. A page behind a login, an internal company network or a paid database is deep web content because a search engine does not index it, not because anyone is hiding anything illegal.

What decides what is "surface"

Google's documentation explains why indexing is not universal. It says Google uses crawlers to find pages for its index, that there is no central registry of all web pages, and that indexing is not guaranteed: not every page Google processes will be indexed. Some pages are disallowed from crawling by the site owner, and others are not accessible without logging in. So the surface web is not the whole open web but the part a search engine has found and chosen to index, and the boundary moves.

How the dark web is reached

CRS says the dark web can be reached through decentralized, anonymized networks such as Tor, short for The Onion Router, or I2P. It reports that Tor was originally created by the U.S. Naval Research Laboratory and first released as the Onion Routing project in 2002. Tor routes traffic through three relays, encrypted between hops, and the last, the exit relay, is the address that appears as the source of the traffic. The Tor Project's own page states that onion services "can only be accessed over Tor." The page does not use the terms "deep web" or "dark web."

CRS notes that tools such as Tor2web may allow access to Tor-hosted content without installing the Tor software but without Tor's privacy protections. It adds that Tor is often slower than other browsers because traffic passes through several relays.

Use, risk and anonymity

CRS says the dark web "may be used for legitimate purposes as well as to conceal criminal or otherwise malicious activities." Reasons cited for anonymizing online activity include free speech, anticensorship, reaching blocked content, business confidentiality, discussing sensitive health or victimization topics, and journalists communicating with whistleblowers. CRS also says criminal activity occurs on the deep and dark web as it does on the surface web, and that much illicit activity sits on darknet markets.

CRS stresses that anonymity is not guaranteed: researchers, security experts and law enforcement are constantly developing ways to deanonymize hidden services or users, and some internet service providers can see that a person is using Tor, though not the content. For law enforcement, CRS describes multiagency efforts such as the Joint Criminal and Opioid Darknet Enforcement team, led by the FBI.

How large is each layer

There is no reliable count. The 2022 CRS overview says data on the magnitude of the deep and dark web are unclear, but data on Tor users exist: according to Tor Project metrics, the United States had a mean of 545,028 daily Tor users across the first six months of 2022, or 21.3% of the worldwide mean. The earlier CRS report said that between March 2016 and March 2017 there were generally 50,000 to 60,000 hidden services, or unique .onion addresses, per day, and that early estimates of the deep web at 4,000 to 5,000 times the surface web were overtaken by how information is accessed. These are dated figures, not current totals.

A short way to remember it

  • Can a search engine index it? Surface web.
  • Is it unindexed because it is private, paywalled or behind a login? Deep web.
  • Is it deliberately hidden and reached with Tor or similar software? Dark web.

Take this further with Oakgen Agent

Agent can research with web search, explain the layers, write a quiz and create a downloadable docx or pdf handout.

Coming soon: Browsing hidden or Tor-only sites from inside Agent is not available.

Looking for a recap of one specific YouTube video? The YouTube Summarizer does that today.

Open this task in Oakgen Agent

The prompt is filled in for you. Nothing is sent until you press send.

Open in Oakgen Agent

Frequently asked questions

What are the levels of the internet?

The Congressional Research Service describes the web as having layers: the surface web of content indexed by traditional search engines, the deep web of unindexed content, and, within the deep web, the dark web of intentionally concealed content that generally needs special software. Numbered lists with more than three levels circulate online, but none of the sources opened for this guide use them.

Is the deep web the same as the dark web?

No. CRS says the deep web is content that has not been indexed, such as private intranets and commercial databases like Westlaw that often require authentication. The dark web is the segment of the deep web that has been intentionally hidden and generally cannot be accessed without special software such as Tor. Most deep web content is ordinary, for example anything behind a login.

How do people reach the dark web?

CRS says it can be reached through decentralized, anonymized networks including Tor and I2P. The Tor Project says onion services can only be accessed over Tor. CRS also notes that Tor was originally created by the U.S. Naval Research Laboratory, first released as the Onion Routing project in 2002.

Is the dark web anonymous and illegal?

CRS states that anonymity on the dark web is not guaranteed and that law enforcement uses specially designed techniques to bypass anonymity protections. It also says anonymizing services have been used for legal and illegal activities, from protecting free speech and communications with journalists to selling illicit goods.

How big is the deep web?

Nobody has a reliable figure. A 2022 CRS overview says data on the magnitude of the deep and dark web and how they relate to the surface web are unclear. An earlier CRS report noted that early estimates putting it at 4,000 to 5,000 times the surface web had been overtaken by changes in how information is presented, and that measuring it is almost impossible.

Sources

  1. Congressional Research Service, The Dark Web: An Overview (IF12172), 22 July 2022 (read 2026-11-19)
  2. Congressional Research Service, Dark Web (R44101), 2015 to 2017 versions, via EveryCRSReport (read 2026-11-19)
  3. Tor Project, Onion Services (read 2026-11-19)
  4. Google Search Central, How Google Search works (read 2026-11-19)
Share