Best Dark Web Search Engines for 2026

What a Dark Web Search Engine Can and Cannot Do

A dark web search engine indexes .onion sites—services that run on the Tor network and hide the server's IP address[5][6]. Unlike Google or Bing, these crawlers reach only a small fraction of what exists: many onion services are invite-only, protected by authentication, or live for hours before disappearing. The largest publicly claimed index is Haystak's self-reported 1.5 billion pages, though no independent audit confirms that figure[4]. Torch launched in 2013 with around 130,000 resources[3], and Ahmia began in 2010 because no other search engine covered hidden services at the time[7]. Even today, no crawler sees everything.

Clearnet mirrors versus native onion crawlers

You will encounter two types of tool. Clearnet search engines with .onion mirrors—DuckDuckGo's onion address, Ahmia's clearnet interface—let you query from a regular browser or from inside Tor Browser, but the index itself is built by a crawler that walks onion links. Native onion crawlers like Torch run entirely within the Tor network, discovering new services by following hyperlinks between .onion pages. Both approaches miss private forums, paywalled markets, and any service that blocks automated requests or requires a login before showing content.

Known address versus discovery

If you already have an onion URL—copied from a trusted source or a vetted directory like The Hidden Wiki—you can paste it directly into Tor Browser's address bar; no search engine is needed. Search engines become useful when you want to discover new services by keyword: typing 'whistleblowing platform' into Ahmia will surface SecureDrop instances and similar tools[6], but it will not reveal a private marketplace that never advertises its name or a single-page blog that no crawler has yet visited. Treat search results as a starting shortlist, not a complete map.

How to Evaluate a Dark Web Search Engine Before You Use It

Before you trust any dark web search engine, check five concrete criteria: whether it publishes its indexing policy, filters abuse material, maintains uptime history, logs your queries, and whether it is reachable only via Tor or also through the clearnet.

Transparency and abuse filtering

A reputable engine explains what it indexes and what it blocks. Ahmia publishes its source code on GitHub—divided into an Elasticsearch index, an onion crawler, and the search website—and maintains a public blacklist of abuse material[1]. That openness lets you verify how the crawler behaves and what content it excludes. Engines that promise 'everything' or boast billions of unverified pages often index spam, scam phishing clones, and illegal material without moderation; the widely repeated claim that Torch indexes over a billion pages cannot be traced to any verifiable source[4]. If an engine does not state its filtering policy, assume it shows whatever the crawler finds.

Query logging and access method

Check whether the service logs your search terms or IP address. Ahmia's clearnet interface at ahmia.fi does not require Tor Browser, which means your ISP sees that you visited ahmia.fi but not which onion links you clicked afterward; the onion mirror keeps all traffic inside Tor[5]. Engines that demand registration before showing results—asking for an email address or username—create a permanent link between your identity and your queries. Avoid them: legitimate onion search engines work without accounts.

Uptime and reachability

Onion services can vanish without notice. An engine with years of documented operation—Ahmia since 2010, Torch since October 2013—is more likely to remain available than a new service with no history[7][3]. Test the .onion address in Tor Browser before you rely on it: if the page loads quickly and returns results for a common term like 'forum', the service is active. If it times out or shows a blank page, move to the next option on your list.

Quick evaluation checklist

Use this table as a starting point: visit each engine's .onion address in Tor Browser, run a test query, and confirm the service loads before you bookmark it. Engines that require JavaScript, ask for payment, or redirect you to clearnet registration pages introduce unnecessary risk—skip them and choose a simpler alternative.

Ahmia: The Safest Starting Point for Most Users

Ahmia is an open-source search engine built specifically for Tor hidden services, with its code freely available on GitHub and divided into an Elasticsearch index, an onion crawler, and the search website. Unlike engines that index everything they find, Ahmia explicitly does not allow abuse material in its index and maintains a public service blacklist, removing reported abuse material as soon as possible[1]. That policy makes it the safest first choice for journalists, researchers, and anyone new to onion services who wants a curated experience rather than raw, unfiltered results.

Clearnet and onion access

You can reach Ahmia two ways: through its clearnet interface at ahmia.fi or via its .onion address inside Tor Browser. The clearnet mirror lets you search without installing Tor, but your ISP will see that you visited ahmia.fi; clicking an onion link from the results page will still require Tor Browser to open[5]. The onion mirror keeps your entire session—query, results, and clicks—inside the Tor network, so no observer outside Tor sees which terms you searched or which services you visited. If you already run Tor Browser, use the onion address; if you want to preview results before committing to Tor, start with the clearnet site.

Who built it and why

Ahmia development began in 2010 as a side project when its creator registered ahmia.fi, because at that time there were no search engines for searching hidden service web content[7]. The Tor Project has since backed the engine, and Ahmia's stated mission is to create the premier search engine for services residing on the Tor anonymity network and to share meaningful statistics, insights, and news about the Tor network[2]. That institutional support means the project follows the Tor community's norms: transparency about what it indexes, active moderation of illegal content, and a focus on legitimate use cases like metadata-free chat, safer interaction between journalists and their sources through tools such as SecureDrop, and more secure ways to reach popular websites[6].

What Ahmia filters and what it keeps

Ahmia's public blacklist removes child sexual abuse material and other abuse content as soon as the team receives a report[1]. The engine does not promise to block every scam or phishing clone—those appear and disappear faster than any manual review can track—but it does prevent the most harmful material from showing up in search results. If you are a first-time user or work in a role where accidental exposure to illegal content creates legal or ethical risk, Ahmia's filtering policy is the strongest available among onion search engines. Engines like Torch do not state a filtering policy, and Haystak's moderation practices remain undocumented[4].

When to choose Ahmia over alternatives

Use Ahmia when you need a vetted shortlist of onion services and want to avoid unmoderated indexes. It works well for discovering whistleblowing platforms, privacy-focused forums, and mirrors of clearnet sites that offer onion addresses for censorship resistance[6]. Ahmia will not surface every private marketplace or invite-only forum—those services either block crawlers or never publish links that a search engine can follow—but it will show you the subset of onion services that their operators intend to be publicly discoverable and that pass Ahmia's abuse filter. For broader, unfiltered coverage, add Torch to your workflow; for a larger self-reported index, try Haystak[3][4]. Start with Ahmia, then expand only if its results do not meet your need.

Torch: The Oldest Index, But Use It Carefully

Torch launched on October 10, 2013, and at that time indexed almost 130,000 Tor resources[3]. It remains one of the longest-running onion search engines, and you will often see it recommended in forums and directories like The Hidden Wiki. The widely repeated claim that Torch indexes over a billion pages cannot be traced to any verifiable source—the only billion-scale figure traceable to an operator is Haystak's self-reported 1.5 billion pages, none of which has been independently verified[4]. What matters more than the claimed size is Torch's indexing approach: unlike Ahmia, Torch does not state a filtering policy for abuse material or illegal content, so its results include whatever the crawler finds.

What you will see in Torch results

Torch returns a broader, unfiltered set of onion services than Ahmia. If you search for a niche forum—say, a privacy-focused discussion board about encryption tools—Torch may surface older or less-maintained services that Ahmia's moderation has excluded or that never submitted to Ahmia's index. That breadth is useful when Ahmia returns zero results for an obscure term. The trade-off is risk: the same query can also return scam marketplaces, phishing clones of legitimate sites, and links to illegal services. Torch does not warn you which category a result falls into, so every link requires the same caution you would apply to an unknown URL posted in an anonymous forum.

Practical safety steps when using Torch

Keep JavaScript disabled in Tor Browser's security settings—set the slider to 'Safest'—before you open any Torch result[5]. Many scam pages and exploit kits rely on JavaScript to fingerprint your browser or deliver malware; disabling it blocks those techniques at the cost of breaking some legitimate site features. Do not click a link unless you recognise the service name or can verify it through a second, trusted source. If a result promises a well-known marketplace or forum, cross-check the onion address against a vetted directory or community recommendation before you visit. Scammers clone popular service names and buy similar-looking domains; one character difference in a 56-character .onion address is easy to miss.

When Torch adds value and when it does not

Use Torch after you have tried Ahmia and need wider coverage, or when you are researching the full landscape of onion services for a project that requires unfiltered data. Do not use Torch as your first search engine if you are new to the dark web or unfamiliar with identifying scam indicators—phishing pages, fake escrow services, and malware-hosting sites appear in results without any flag. Torch works best as a secondary tool for experienced users who already know how to verify onion addresses, recognise common scam patterns, and navigate Tor Browser's security settings. If you accidentally encounter disturbing or illegal content, close the tab immediately and do not return to that address; Ahmia's filtering exists precisely to prevent that scenario[1].

DuckDuckGo on Tor: A Clearnet Engine with an Onion Mirror

DuckDuckGo is not a dark web crawler and does not index .onion sites the way Ahmia or Torch do[3]. It is a clearnet search engine—one that indexes the public web—that offers an onion service mirror so you can search the regular internet privately from inside Tor Browser[5]. When you visit DuckDuckGo's .onion address, you run the same queries you would on duckduckgo.com, but all traffic between your browser and the search engine stays inside the Tor network, hidden from your ISP and any observer monitoring your connection. That setup is useful when you want to search for clearnet information—news articles, software documentation, forum threads on Reddit—without leaving Tor or revealing your IP address to DuckDuckGo's servers.

What DuckDuckGo will and will not find

DuckDuckGo's index covers websites with ordinary domain names (.com, .org, .co.uk) and will not discover onion services unless those services also operate a clearnet mirror that DuckDuckGo's crawler can reach[6]. If you search for 'Tor whistleblowing platform', you may see results about SecureDrop's clearnet documentation, but you will not see a list of individual .onion addresses the way you would in Ahmia. DuckDuckGo does filter search results for policy violations and does not log your queries or IP address under its stated privacy policy, but it remains a tool for searching the public web, not for exploring onion services[1]. Use it when you need to look up a guide, verify a claim, or find contact information for a clearnet organisation while staying inside Tor; switch to Ahmia or Torch when you need to discover onion-only resources.

When DuckDuckGo's onion mirror makes sense

Choose DuckDuckGo's onion service if you already browse the clearnet through Tor Browser and want a familiar search interface that respects your privacy[6]. It works well for fact-checking an article, searching technical documentation, or finding a clearnet forum discussion about a specific onion service before you visit that service. Because DuckDuckGo does not index .onion addresses, it will not replace a dedicated onion search engine in your workflow—treat it as a complement, not a substitute. If you need to search both the clearnet and the dark web during the same session, bookmark DuckDuckGo's onion address alongside Ahmia's; that way you can switch between them without leaving Tor or exposing your queries to a clearnet DNS resolver.

Other Engines Worth Knowing: Haystak, Recon, and Kilos

Three other onion search engines appear frequently in dark web directories and forum recommendations: Haystak, Recon, and Kilos. Each serves a more specialised audience than Ahmia or Torch, and all three carry higher risk of exposure to illegal content because their filtering policies—if they exist—remain undocumented or explicitly target marketplaces and vendor data. The widely repeated claim that Haystak indexes 1.5 billion pages is the only billion-scale figure traceable to an operator, but none of these self-reported index sizes has been independently verified[4]. Treat these engines as research tools for experienced users, not as safe entry points for newcomers.

Haystak: Aggressive indexing and premium claims

Haystak markets itself as the largest onion search engine by page count and offers a free tier with basic search alongside a premium subscription that promises faster results and access to additional filters[4]. The engine does not state a public abuse-material policy comparable to Ahmia's blacklist, so you should expect unfiltered results that include scam sites, phishing clones, and links to illegal services[1]. Haystak's premium model is unusual among onion search engines—most operate as free, community-funded projects—and you cannot verify whether the paid tier delivers meaningfully better coverage without subscribing. Use Haystak only after you have exhausted Ahmia and Torch, and only if you need the broadest possible index for a specific research task that justifies the exposure risk.

Recon and Kilos: Marketplace and vendor search

Recon and Kilos focus explicitly on indexing dark web marketplaces and vendor listings rather than general onion services[6]. Recon aggregates vendor profiles, product categories, and feedback from multiple marketplaces, while Kilos functions as a market search tool that lets you query listings across several platforms simultaneously. Both engines assume you already understand marketplace mechanics, escrow systems, and the legal risks of browsing commerce platforms that host illegal goods. If you are researching the structure of dark web markets for journalism, academic study, or threat intelligence, these tools provide targeted data that general engines do not surface. If you are new to Tor or unfamiliar with identifying scam vendors and phishing markets, do not use Recon or Kilos—accidental clicks can expose you to law enforcement monitoring, malware, or financial fraud[8].

Honeypot and scam risk across all engines

Any onion search engine can be a honeypot operated by law enforcement, a scam designed to harvest credentials, or a malware distribution point disguised as a legitimate service[8]. The FBI's Network Investigative Technique deployed on the Playpen hidden service in 2015 demonstrated that code added to an onion service can force users' computers to transmit identifying information such as IP address, host name, and MAC address back to a government-controlled server[8]. No single search engine can be verified as safe through its interface alone. Before you rely on Haystak, Recon, Kilos, or any engine not covered in this guide, cross-check its onion address against multiple trusted sources—community forums with established reputations, vetted directories, or direct recommendations from users whose operational security you trust. If an address appears in only one location or comes with aggressive marketing claims, treat it as unverified until you see corroboration. Keep Tor Browser's security level at 'Safest', never enter personal information into a search engine, and remember that even a legitimate engine can return links to hostile sites[5].

Typical Mistakes That Lead to Malware, Scams, or Legal Trouble

Most security failures on the dark web stem from five preventable errors. Each mistake exposes you to malware, financial loss, or legal risk that proper habits would avoid. Below are the most common missteps and the safer alternative for each.

Clicking .onion links from random Reddit threads

You find a thread titled 'Best onion markets 2026' on a general subreddit and click the first .onion address someone posted. That link can point to a phishing clone, a malware host, or a honeypot operated by law enforcement[8]. Reddit allows anyone to post, and scammers seed threads with fake addresses that look one character different from legitimate services. The safer approach: never visit an .onion address unless you have verified it through at least two independent, trusted sources—a vetted directory with a known maintainer, a long-established forum with moderation, or direct confirmation from the service's clearnet presence. Dark Web Link Reddit: Finding Reliable Sources explains how to cross-check addresses before you click.

Enabling JavaScript on unknown onion sites

Tor Browser ships with a security slider that disables JavaScript when set to 'Safest', but many users leave it at the default 'Standard' level because some sites break without scripts[5]. Visiting an unknown onion service with JavaScript enabled lets that site run code in your browser, which can fingerprint your system, exploit browser vulnerabilities, or force your computer to transmit your real IP address to a remote server[8]. The FBI's Network Investigative Technique added to the Playpen hidden service in 2015 used exactly this method to capture IP addresses, host names, and MAC addresses from users who had JavaScript enabled[8]. Set the security slider to 'Safest' before you open any search result, especially from unfiltered engines like Torch. You will lose some site functionality, but you block the primary vector for browser-based exploits.

Trusting search results as vetted or safe

Ahmia filters abuse material and maintains a public blacklist, but Torch, Haystak, Recon, and Kilos do not state comparable policies[1][4]. A search result in any of these engines can lead to a scam marketplace, a phishing page that harvests login credentials, or a site hosting illegal content. No onion search engine reviews every indexed page for safety, and even Ahmia's filtering relies on user reports rather than real-time inspection. Treat every link as hostile until you verify it. Look for corroboration from community forums, check the onion address character by character against a known-good source, and never assume that appearing in search results means a site is legitimate or legal.

Using a VPN instead of Tor for anonymity

A VPN hides your IP address from the websites you visit, but your VPN provider sees every site you connect to and can log that activity or hand it to law enforcement under legal compulsion. Onion services are only accessible through the Tor network, which encrypts traffic in layers and routes it through multiple relays so no single point—including the Tor Project—knows both your identity and your destination[5][6]. If you try to reach an .onion address through a VPN without Tor Browser, you will fail; the .onion domain does not resolve in ordinary DNS. If you use a VPN and Tor together, the VPN sees that you are connecting to Tor but cannot see which onion services you visit, while Tor still provides the anonymity layer. For most users, Tor Browser alone is sufficient and simpler. Add a VPN only if your threat model includes an adversary who can monitor your internet connection and you want to hide the fact that you use Tor at all—not as a replacement for Tor.

Assuming all dark web content is illegal or all legal

The Tor Project's own materials state that the 'dark web' label is often used negatively in the context of encouraging encryption backdoors, and that onion services host metadata-free chat, file sharing, safer journalist–source interaction through SecureDrop, and more secure access to popular clearnet sites like Facebook[9][6]. Visiting an onion search engine or browsing .onion addresses is legal in the UK and most jurisdictions. What matters is what you do once you arrive: viewing, purchasing, or distributing illegal material—child abuse images, stolen data, controlled substances—is a crime regardless of the technology you use to access it. Conversely, not every onion service is a marketplace or forum for illegal activity; many host privacy tools, academic research, and whistleblowing platforms. Do not let the 'dark web' label make you reckless because you assume everything is forbidden, and do not assume you are safe from prosecution simply because you used Tor.

How to Verify That an Onion Link Is Still Active and Safe

Before you visit any .onion address from a search result, verify it against at least two independent sources and check that the service is still reachable. Onion services can disappear without notice—operators shut them down, law enforcement seizes them, or hosting infrastructure fails—and a link that worked last month may now point to nothing, a phishing clone, or a honeypot. No search engine updates its index in real time, so you must confirm each address yourself before you click.

Step one: check the engine's own status page or mirror list. Ahmia publishes its source code and development updates on GitHub, and the Tor Project maintains a list of official onion services for its own properties[5]. If the search engine you are using provides a clearnet status page or a known mirror list, compare the .onion address character by character. A single wrong letter can redirect you to a phishing site. If the engine offers no official list, move to step two.

Step two: cross-check the address in a vetted directory. The Hidden Wiki: Your Guide to the Dark Web's Directory and similar community-maintained directories list onion services with descriptions and update dates. Look for the same address in at least two directories that have different maintainers. If the address appears in only one place, or if the directories show conflicting addresses for the same service name, treat it as unverified. Do not visit until you find corroboration from a forum with established moderation or a direct announcement from the service's clearnet presence.

Step three: set Tor Browser's security slider to 'Safest' before you test the link. Open Tor Browser, click the shield icon in the address bar, and move the slider to 'Safest'. This setting disables JavaScript, which blocks the primary method hostile sites use to fingerprint your system or exploit browser vulnerabilities[5][8]. The FBI's Network Investigative Technique deployed on Playpen in 2015 used JavaScript to force users' computers to transmit IP addresses and MAC addresses to a government server; 'Safest' mode would have blocked that code from running[8]. Some onion services will not load fully without scripts, but you can still see whether the address resolves and whether the landing page matches the service description.

Step four: visit the address and check for a timeout or error. If Tor Browser shows 'Unable to connect' or 'Onionsite not found', the service is offline or the address is incorrect. Wait five minutes and try again; onion routing can be slow, and temporary network congestion sometimes causes false negatives. If the error persists after three attempts over an hour, the service is likely gone. Do not assume it will return—many onion services close permanently without announcement.

Conditional scenario: a link worked last month but now times out. You bookmarked an onion search engine in January, and in February the address no longer loads. First, check whether the engine announced a new mirror or address change on its clearnet site or in a community forum you trust. Ahmia, for example, operates both ahmia.fi and an .onion mirror; if one goes down, the other may still list the current address[1]. Second, search for recent mentions of the service name in forums or on Reddit to see whether other users report the same problem or share a working alternative. If you find no updates after 48 hours, treat the service as defunct and remove the bookmark. Do not click on 'replacement' addresses posted by anonymous users without verifying them through the four steps above—scammers post fake mirrors immediately after a popular service goes offline.

Frequently Asked Questions About Dark Web Search Engines

Which search engine is used in the dark web?

Ahmia, Torch, and DuckDuckGo are the most widely used search engines for onion services. Ahmia is an open-source engine that indexes Tor hidden services and filters abuse material from its results, with development that began in 2010 when no other search engines existed for hidden service content[1][7]. Torch launched in October 2013 and was one of the first large-scale onion search engines, though its widely repeated claim of indexing over a billion pages cannot be traced to any verifiable source[3][4]. DuckDuckGo operates a clearnet search engine that you can access through Tor Browser, but it does not index .onion addresses—it returns results from the surface web while preserving your privacy through Tor's encrypted routing[5]. If you want to search specifically for onion services, use Ahmia or Torch; if you want general web search without tracking, use DuckDuckGo through Tor Browser.

Can the FBI track Tor?

The FBI can identify Tor users under specific conditions, but it cannot monitor all Tor traffic or break the network's encryption by default. In 2015 the FBI deployed a Network Investigative Technique on the Playpen hidden service, adding computer code that caused users' browsers to transmit their real IP address, host name, operating system username, and MAC address to a government server[8]. This technique worked only because users visited the compromised site with JavaScript enabled; Tor Browser's 'Safest' security level disables JavaScript and would have blocked the code from running[5][8]. The Ninth Circuit later held that the seized information—mainly the IP address—helped identify a user but provided no information about the computer's or user's precise physical location at the time[10]. The FBI cannot see your Tor traffic if you keep your browser updated, leave the security slider at 'Safest', and avoid downloading files or enabling plugins that can leak your identity. Tor's onion routing encrypts your traffic in layers and routes it through multiple relays so no single point knows both your identity and your destination[5]. Law enforcement can compromise individual onion services or exploit user mistakes, but the Tor network itself remains resistant to mass surveillance.

Is Tor Browser 100% safe?

No browser is 100% safe, and Tor Browser is no exception. Tor hides your IP address and encrypts your traffic through multiple relays, but it cannot protect you from malware, phishing, or mistakes you make after the connection is established[5][6]. If you download a file from an onion site and open it outside Tor Browser, that file can contain malware that calls home with your real IP address. If you enable JavaScript on an unknown site, hostile code can fingerprint your system or exploit browser vulnerabilities, as the FBI's Network Investigative Technique demonstrated in 2015[8]. If you enter personal information into a search engine or log into an account that ties to your real identity, you remove the anonymity Tor provides. Tor Browser gives you strong privacy and anonymity if you follow safe practices: keep the security slider at 'Safest', never enable plugins, do not download files unless you can verify them, and treat every onion link as hostile until you confirm it through independent sources. Onion services guarantee end-to-end encryption and make it difficult for adversaries to censor them or identify their operators, but they do not guarantee that the content inside is safe or legal[5][11].

What are some good search engines other than Google?

DuckDuckGo, Startpage, and Brave Search are privacy-focused alternatives to Google for clearnet searching, and all three work through Tor Browser without requiring JavaScript. DuckDuckGo does not track your searches or build a profile on you, and it returns results from Bing and its own index. Startpage anonymises your queries and returns Google results without Google seeing your IP address or search history. Brave Search uses its own independent index and does not track you or sell your data. None of these engines index .onion addresses, so you will not find onion services in their results[6]. If you want to search onion content specifically, use Ahmia or Torch as described in the main sections of this guide. If you want general web search with privacy, any of the three clearnet engines above will serve you well through Tor Browser, and DuckDuckGo is the default search provider in Tor Browser because it requires no additional configuration[5].

Typical Mistakes and Misconceptions

Assuming every 'dark web search engine' actually searches onion sites

Many lists mix general onion directories with true search engines, so you may install a tool that never crawls .onion content. DuckDuckGo, for example, works well through Tor Browser but returns surface web results and does not index .onion addresses[6]. Ahmia and Torch, by contrast, are built specifically to index Tor hidden services[3]. Before you rely on any engine, open its results and check whether the links end in .onion or in ordinary domains.

Believing the 'billion pages indexed' claims without checking the source

The widely repeated figure that Torch indexes over a billion pages has no verifiable origin, and the only billion-scale claim traceable to an operator is Haystak's self-reported 1.5 billion pages[4]. No independent audit confirms either number, so treat them as marketing rather than fact. A more reliable signal is whether the engine publishes its crawler code and index methodology, as Ahmia does on GitHub. When a number cannot be traced, assume the actual index is smaller.

Thinking a dark web search engine filters illegal content by default

Only some engines maintain an abuse blacklist, and absence of filtering is not the same as safety. Ahmia explicitly excludes abuse material and removes reported content from its public index as soon as possible[1]. Torch and Haystak publish no confirmed filtering policy, which means their results can surface marketplaces and other illegal services without warning[4]. If you want a curated starting point, choose an engine with a documented blacklist and still verify every link yourself.

Confusing the 'deep web' with the 'dark web'

The deep web is simply content not indexed by ordinary search engines, often hidden behind passwords or paywalls, and the term predates Tor's popularity[9]. The dark web refers specifically to onion services on the Tor network, which use the special-use .onion top-level domain and require Tor Browser to reach[6]. A database behind your bank login is deep web; an onion forum is dark web. Mixing the terms leads you to expect search engines to find things they were never designed to index.

Assuming onion search engines are immune to law enforcement

Onion services hide server location and encrypt traffic, but that does not protect you from malicious code on the sites themselves[5]. In 2015 the FBI added a Network Investigative Technique to the Playpen hidden service, causing visitors' browsers to transmit IP addresses, host names, and MAC addresses to a government server[8]. The technique relied on JavaScript, which is why Tor Browser's 'Safest' security level matters before you click any search result[8]. An engine can point you to a compromised site, and no index filters for that risk.

Expecting search results to stay current without manual checks

Onion services disappear without notice when operators shut them down, law enforcement seizes them, or hosting fails, and no engine updates its index in real time. A link that worked last month may now lead to a phishing clone or a honeypot, so you must verify addresses against at least two independent sources before clicking. Ahmia's open-source crawler and public blacklist make its current index easier to audit, but even that does not guarantee a given service is still live[1]. Treat every result as a lead to confirm, not a destination to trust.

Conclusions

• Start with Ahmia: it publishes its crawler code on GitHub and maintains a public abuse blacklist, so you can audit what it indexes and what it removes[1].

• Treat every result as a lead to confirm, not a destination to trust. Onion services disappear without notice, and no engine updates its index in real time.

• Check whether results end in .onion before you rely on any engine. DuckDuckGo works well through Tor Browser but returns surface web results and does not index onion addresses[6].

• Set Tor Browser's security level to 'Safest' before clicking any search result. The FBI's 2015 Playpen operation relied on JavaScript to transmit visitors' IP addresses and MAC addresses to a government server[8].

• Ignore index-size claims you cannot trace. Torch's 'billion pages' figure has no verifiable source, and Haystak's self-reported 1.5 billion pages have never been independently confirmed[4].

Your first step is to open Ahmia, run a query, and confirm the results end in .onion rather than ordinary domains. For a broader starting point, see List of Onion Sites: Explore the Dark Web Safely.