Deep Web: definition, examples, and difference from the dark web

Published April 17, 2026 · Updated April 17, 2026 · About 12 minutes read

The deep web is probably the most misunderstood concept on the Internet. Constantly confused with the dark web in mainstream media, presented as an obscure zone populated by criminals, it actually refers to a far more mundane notion: all online content that is not indexed by Google, Bing, or other standard search engines. Your Gmail inbox, your bank account, your company's intranet, university library catalogs, your Google Drive — all are part of the deep web. This guide clearly explains what the deep web is, its actual size, what it concretely contains, and above all what it is not — notably the equivalent of the dark web.

What is the deep web?

The deep web (also called the hidden web or invisible web) refers to all pages and online content that are not indexed by standard search engines such as Google, Bing, DuckDuckGo, or Qwant. The term was introduced in 2001 by Michael K. Bergman in a research article entitled The Deep Web: Surfacing Hidden Value, which was the first attempt to estimate the size of this non-indexed part of the Internet.

The definition fits in one sentence: if a page does not appear in Google results, it is part of the deep web. This includes everything that requires authentication, a specific database query, restricted access, or simply the absence of incoming links allowing a crawler to discover it. The deep web is therefore not a physical place or a separate network: it is a property of certain pages (their invisibility to search engine crawlers).

Actual size of the deep web

Historical estimates of the size of the deep web date essentially from Bergman's foundational article in 2001. He estimated the deep web at approximately 500 times the size of the surface web at the time, an order of magnitude likely exceeded today by the explosion of cloud services, private social networks, and corporate databases.

Modern estimates vary according to methodology and what is chosen to count (HTML pages alone, or also multimedia files, emails, databases), but all converge on the same order of magnitude: the deep web represents approximately 90 to 96% of the total Internet content. The surface web (pages indexed by Google) is therefore only the tip of a much larger iceberg.

To put these figures in perspective: Google indexes approximately 60 to 100 billion web pages. The deep web probably contains several tens of trillions, if we count private emails, instant messages, cloud files, corporate documents, medical databases, and government archives accessible only on request.

Concrete examples of deep web content

The best way to understand the deep web is to list what it contains. Most of this content is perfectly familiar to you.

Your personal accounts

Your Gmail, Outlook, ProtonMail, Yahoo inbox
Your online banking
Your social media accounts (private zones, direct messages)
Your Google Drive, Dropbox, iCloud, OneDrive
Your Netflix, Spotify, Amazon Prime, Disney+ subscriptions
Your online games and their match history

Public and institutional services

Government tax portals (your personal area, not the public site)
Health insurance portals (your reimbursements, your prescriptions)
Social benefit portals (personal files)
School and university digital learning environments
Library catalogs accessible via specific queries

Specialized databases

PubMed, ScienceDirect, JSTOR (paid or access-restricted scientific publications)
Municipal and regional archives
Commercial registries, property records, cadastre
National business databases

Corporate content

Intranets (SharePoint, Confluence, internal Slack, Teams)
CRM (Salesforce, HubSpot with customer data)
ERP (SAP, Oracle)
Project management tools (Jira, private Trello, Asana)
Corporate document repositories

This list shows the breadth of the deep web and its perfectly ordinary nature. None of these pages have any connection with the dark web or criminal activities — they are simply protected by authentication for obvious reasons of privacy and security.

Deep web vs dark web: the common confusion

The confusion between the deep web and the dark web is the most widespread error in mainstream articles on the subject. The two terms refer to radically different realities. For a thorough comparison, see our dedicated article dark web vs deep web.

Criterion	Deep web	Dark web
Nature	Content not indexed by Google	Sites accessible only via Tor/I2P
Access	Regular browser + credentials	Dedicated software (Tor Browser)
Size	90-96% of the Internet	Less than 0.01% of the Internet
Anonymity	None (named authentication)	Strong (onion routing)
Examples	Gmail, intranet, bank account	BBC .onion, SecureDrop, marketplaces
Legality	Completely legal and mundane	Legal as such, varied content

The dark web is therefore an extremely minority sub-part of the deep web: all .onion pages are part of the deep web (they are not indexed by Google), but the vast majority of the deep web has nothing to do with Tor or anonymity. Saying "the deep web is dangerous" is the same as saying your Gmail inbox is dangerous.

Deep web vs surface web (clear web)

The surface web (or clear web) is the complement of the deep web: everything that is indexable and indexed by standard search engines. This is the part of the Internet you browse when you type a query into Google: news sites, blogs, Wikipedia, public online stores, institutional sites, social networks (public zones).

The boundary between the surface web and the deep web is fluid and depends on publisher policies: an article that is free but behind a basic paywall can shift from the surface web to the deep web if Google loses access. Conversely, an initially private page can become public and indexable with a simple permission change. The criterion is operational (is it in the Google index?) and not technical.

Why Google doesn't index the deep web

Several reasons explain why Google and other search engines do not index the majority of online content.

Authentication required

Google's crawler (Googlebot) does not have your credentials. It cannot log into your Gmail account or your bank account. These pages are therefore invisible to it, not because they are technically hidden but because it does not have the access rights.

Exclusion instructions (robots.txt, noindex)

Website administrators can explicitly forbid indexing via a robots.txt file or meta robots noindex tags. Googlebot respects these instructions. Intranets, staging environments, and internal APIs systematically use these mechanisms.

Dynamically generated content

Many pages only exist when a user makes a specific request: search results on a catalog, a contact form, a personalized document generator. Googlebot does not submit forms and does not discover these pages.

Absence of incoming links

Google discovers pages by following links. A page with no incoming links from an already-indexed site remains invisible, even if it is technically accessible. This is the case for many municipal archive pages, specialized databases, or PDF files hosted without any public reference.

How to access the deep web

You access the deep web every day without realizing it. No special tools are required: an ordinary browser suffices, with the right credentials.

Logging into your email → deep web
Checking your online bank account → deep web
Accessing your company's intranet → deep web
Searching in a university database → deep web
Opening your personal government services area → deep web

To access the dark web (a specific sub-part of the deep web), the approach is different: you need to install Tor Browser and use .onion addresses. Our step-by-step guide how to go on the dark web details the complete procedure.

Deep web and privacy

The deep web is fundamentally linked to privacy protection. If all online content were automatically indexable by Google, it would be a privacy catastrophe: your emails, your bank statements, your medical records would appear in search results. The technical mechanisms that create the deep web (authentication, restricted access, dynamic content) are precisely what protects your digital privacy.

To strengthen this protection against advertising tracking and commercial or governmental surveillance, more tech-savvy users turn to tools like Tor, VPNs, encrypted messaging (ProtonMail, Signal), and amnesic operating systems (Tails OS). See our privacy tools category for available solutions.

Myths about the deep web

Several misconceptions circulate about the deep web, often derived from confusion with the dark web. For a systematic fact-check of associated myths, see our pillar 50 dark web myths debunked.

"The deep web is dangerous"

False. The deep web is perfectly mundane. Your Gmail account is part of the deep web and has nothing dangerous about it. The confusion stems from the media's equation of deep web = dark web, which is incorrect.

"The deep web has hidden levels"

False. The multi-level iceberg (level 1, 2, 3, 4, 5, with "Mariana's Web" at the bottom) is a fictional construct that originated on 4chan around 2011. The Internet has no hierarchical layers: each piece of content is either technically accessible or not. For a complete deconstruction, see our article Mariana's Web: myth or reality.

"You need special software for the deep web"

False. An ordinary browser suffices, plus the credentials required to access the private content that concerns you. Tor is only necessary for the dark web, which is a tiny sub-part of the deep web.

FAQ: your questions about the deep web

What exactly is the deep web?

The deep web refers to all online content that is not indexed by search engines like Google or Bing. This includes your emails, your online bank account, corporate intranets, university databases, password-protected dynamic pages, and private streaming services. The deep web represents the vast majority of the Internet, estimated at 90-96% of total content.

Are the deep web and the dark web the same thing?

No. The deep web is all content not indexed by Google (emails, databases, intranets). The dark web is a very small sub-part of the deep web, accessible only through anonymization networks like Tor. Confusing the two is the most common error in mainstream media. For more details, see our guide on the dark web.

Is the deep web illegal or dangerous?

No, absolutely not. The deep web is a normal and massive part of the Internet. Your Gmail account, your online banking, your corporate intranet — all are part of the deep web. This content is not indexed for reasons of privacy or authentication, not to conceal criminal activities.

How do I access the deep web?

You access it every day without realizing it. Every time you log into your email, your bank account, or your company workspace, you are browsing the deep web. No special software is required: an ordinary browser suffices, with the right credentials. This is a major difference from the dark web, which requires Tor Browser.

Does the deep web really represent 96% of the Internet?

The order of magnitude is correct: the web indexable by Google (surface web) represents a small fraction of total online content. Common estimates place the deep web between 90% and 96% of the Internet, depending on whether you count only HTML pages or also databases, private media files, and messages. This figure is often wrongly attributed to the dark web by the media, which is a major error.

Are there search engines for the deep web?

Yes, for certain parts of the deep web. Specialized engines such as the Internet Archive, Base Search, Google Scholar, or university library portals partially index deep web content (archives, theses, scientific publications). But the core of the deep web (your personal data, intranets) remains deliberately inaccessible to search engines for obvious privacy reasons.