Torch is a search engine for .onion sites accessible only through Tor browser, describing itself as the most recognized dark web search engine and the largest index of onion services (1). Torch reports over 100,000 onion sites and over 4 million pages indexed (2), though exact size remains unclear due to conflicting self-reported statistics and lack of independent verification (2) (3).
What Is Torch Search Engine
Torch is a search engine designed specifically for the Tor anonymity network. It indexes .onion sites—hidden services accessible only through Tor Browser. These sites exist on a separate layer of the internet, invisible to standard search engines like Google or Bing.
According to its own description, Torch has grown into the largest index of onion services and is the most recognized search engine in the Tor network, mapping more of the hidden web than any other crawler still in operation (1). This makes it a primary tool for users trying to discover content within the Tor ecosystem.
The Confusion Around Multiple Torch Instances
A significant challenge when researching Torch is that multiple search engines use the Torch name, each reporting different statistics. One instance reports over 100,000 onion sites and over 4 million pages indexed (2). Meanwhile, a second search engine also using the Torch name claims approximately 860,000 pages indexed (2).
This discrepancy creates confusion about which version is the "official" Torch. For you as a researcher, this means you cannot rely on a single set of statistics when evaluating Torch's capabilities. The existence of multiple instances also raises questions about brand authenticity and data quality.
When you encounter a Torch search engine, check its onion address and compare reported index sizes. Different instances may have different coverage, filtering policies, and update frequencies. This fragmentation is common in the dark web ecosystem, where services can be cloned, forked, or imitated without centralized oversight.
For a broader comparison of dark web search tools, see Onion Links Search Engines: Compare and Choose the Right Tool.
How Large Is Torch's Index
Torch's reported index size varies dramatically depending on which instance you examine and when the statistics were published. According to a June 2026 review, one Torch instance reports over 100,000 onion sites and over 4 million pages indexed (2). However, a second search engine using the Torch name claims approximately 860,000 pages indexed (2). This nearly 5× difference in page counts demonstrates the difficulty of verifying dark web search engine claims.
The Billion-Page Claim and Its Contradiction
Torch has also claimed to index over one billion dark net pages, but this claim has not been independently verified (3). More problematically, the service contradicted itself by stating on its onion about page that it indexed only 2 million hidden services (3). This internal inconsistency raises questions about the reliability of all publicly stated figures.
Understanding Pages vs. Sites as Metrics
The discrepancy between "one billion pages" and "2 million hidden services" may be explained by the difference between pages and sites as metrics (3). A page and a site are not the same unit of measurement. If Torch indexed two million hidden services with an average of 500 pages per service, this would equal the claimed billion pages (3).
This distinction matters when you evaluate search engine coverage. A service might index many sites but only their landing pages, resulting in high site counts but low page counts. Conversely, deep crawling of fewer sites produces more pages but fewer unique domains.
Why These Claims Lack Verification
None of Torch's index size claims have been independently verified (3). Dark web search engines operate without third-party audits, transparency reports, or standardized measurement methods. You cannot compare Torch's statistics directly to academic research or other search engines using consistent methodology.
When evaluating Torch or any dark web search engine, treat all self-reported statistics as unverified claims rather than documented facts. Use search quality and result relevance as practical measures instead of relying on headline numbers.
How Torch Compares to Dark Web Crawling Research
Academic research on dark web crawling provides an independent benchmark for evaluating Torch's index size claims. In a longitudinal study of Tor Hidden Services, researchers starting from 4,121 seed addresses reached 24,900 different sites, with 8,381 sites remaining unreachable after multiple attempts (4). This means that even with systematic crawling methods and academic resources, researchers could access only about 75% of discovered addresses.
What This Reveals About Dark Web Indexing Challenges
The study's findings highlight three key obstacles that affect all dark web search engines, including Torch:
High failure rate: More than one-third of discovered sites could not be reached even after multiple attempts, suggesting many .onion addresses are ephemeral, misconfigured, or intentionally restricted.
Limited seed discovery: Starting from 4,121 addresses and reaching 24,900 sites demonstrates that comprehensive discovery requires extensive crawling infrastructure and time.
No complete map exists: If academic researchers with dedicated resources reached only 24,900 sites, Torch's claim of indexing 100,000 sites or 2 million hidden services (2) (3) cannot be independently verified against this baseline.
Why You Cannot Verify Coverage Claims
The research demonstrates that no single entity—academic or commercial—has a complete map of the dark web. When Torch or any search engine claims a specific index size, you have no independent reference point to confirm accuracy. The study's 24,900 reachable sites represent a documented sample, but even this does not establish the total number of active .onion sites at any given time.
For you as a researcher, this means treating all dark web search engine statistics as unverified self-reports. The academic study provides context for the scale and difficulty of indexing work, but it does not validate or refute Torch's specific claims. Instead, evaluate search engines by testing them with known .onion addresses and comparing result quality across multiple tools.
For practical guidance on finding .onion sites using multiple methods, see How to Find Sites on Tor: Search Engines, Directories & Verification.
Torch vs. Alternative Dark Web Search Engines: Documented Comparison
| Search Engine | Index Size Claims | Content Filtering | Interface Features | Verification Status |
|---|---|---|---|---|
| Torch | 100,000+ sites, 4M+ pages (one instance); 860,000 pages (another); 1B pages claimed but contradicted | No filtering documented; returns all indexed results | Basic search box; no advanced operators documented | No independent verification; conflicting self-reports |
| Ahmia | No documented data | Filters illegal content; blocks child abuse material | No documented data | No documented data |
| Haystak | No documented data | No documented data | No documented data | No documented data |
| DuckDuckGo onion | No documented data | No documented data | No documented data | No documented data |
| Not Evil | No documented data | No documented data | No documented data | No documented data |
Q&A
- How do I access a .onion website?
You need Tor Browser to access .onion sites, as these addresses only resolve within the Tor network. Download Tor Browser from the official Tor Project website, install it, then paste the .onion URL directly into the address bar. Standard browsers like Chrome or Firefox cannot reach .onion addresses even with a VPN.
- Is Torch search engine illegal?
Using Torch itself is not illegal in most jurisdictions, as it is simply a search tool for the Tor network. However, the legality depends on what you search for and access through it—many .onion sites host illegal content, and accessing or interacting with such content may violate local laws. Torch does not filter results for legality, so you are responsible for ensuring your actions comply with applicable laws.
- Can I be tracked if I use Tor browser?
Tor Browser significantly reduces tracking by routing traffic through multiple encrypted nodes, but it does not guarantee complete anonymity. You can still be identified through browser fingerprinting, malicious exit nodes, compromised sites, or by revealing personal information. For stronger protection, avoid logging into personal accounts, disable JavaScript on sensitive sites, and never download files or enable plugins that can leak your real IP address.
- Which search engine is best for the dark web?
No single dark web search engine offers verified comprehensive coverage, so the "best" choice depends on your specific needs and willingness to cross-reference results. Torch claims the largest index but has unverified and contradictory statistics (2) (3), while alternatives like Ahmia focus on filtering illegal content. For reliable discovery, use multiple search engines and directories together rather than relying on any single tool.
- How many dark web sites does Torch index?
Torch's reported index size varies dramatically across instances: one claims over 100,000 onion sites and over 4 million pages (2), while another using the Torch name reports approximately 860,000 pages (2). Torch has also claimed over one billion pages but contradicted itself by stating only 2 million hidden services on its about page (3). None of these figures have been independently verified, so treat all statistics as unverified self-reports.
- Are there multiple versions of Torch search engine?
Yes, multiple search engines use the Torch name, each reporting different index sizes and operating on different .onion addresses (2). This creates confusion about which instance is the "official" Torch and makes it impossible to verify brand authenticity or data quality. When you encounter a Torch search engine, compare its onion address and reported statistics to other known instances, and test result quality with known .onion sites rather than relying on the name alone.
Key Takeaways
Torch claims to be the largest dark web search engine, but its actual index size remains unverified — reported statistics range from 860,000 to over 4 million pages across different instances, with no independent confirmation (2).
Multiple search engines use the Torch name, each operating on different .onion addresses with conflicting coverage claims, making it impossible to identify an authoritative version (2).
The billion-page claim contradicts Torch's own documentation, which stated only 2 million hidden services indexed, highlighting reliability issues with all self-reported figures (3).
Academic research found only 24,900 reachable sites from systematic crawling, demonstrating that comprehensive dark web indexing faces inherent technical limitations no search engine has overcome (4).
Treat all dark web search engine statistics as unverified self-reports — evaluate tools by testing them with known .onion addresses and comparing result quality rather than relying on headline numbers.
To see how Torch compares to other dark web search tools with documented features, explore Onion Links Search Engines: Compare and Choose the Right Tool.
References
[1] About Us — Torch
[2] Torch Search Engine: What It Indexes and What It Misses
[3] Does Torch, a darknet search engine, really index 1 billion hidden services?
Where to look next
These services are useful starting points for further research. Tools & services.