Many alternative search engines serve results supplied by larger providers rather than crawling the web themselves. Google, Microsoft, Yandex, Baidu, Brave, Mojeek, Seznam and Ahrefs are among the organizations operating substantial web-crawling infrastructure. DuckDuckGo states that it runs its own crawler and indexes while sourcing traditional links largely from Bing. Being indexed by a provider is separate from being ranked, served or cited by the products that license its results, and no provider guarantees the second from the first.
Key Takeaways
– Index provenance, licensed results and blended retrieval are distinct arrangements, and several engines sit across more than one
– Bing visibility can support discovery across services using Microsoft infrastructure, without guaranteeing position or inclusion in any of them
– Google states that pages must be indexed and snippet-eligible for AI features, and that “indexing and serving isn’t guaranteed”
– OpenAI recommends allowing OAI-SearchBot to help a site appear in ChatGPT search, and documents no citation guarantee
– Brave reports removing its last Bing API calls in April 2023, a self-reported figure with no independent audit
– Provider relationships change, so verify current arrangements rather than trusting any roundup including this one
Prepared by the SearchEngineInfo Editorial Team. Editorial review completed 16 October 2026. Material claims were assessed against primary provider documentation, independent reporting and available market-share data, and unverified or disputed information is identified in the text. Next scheduled review: 16 January 2027. Provider relationships, crawler policies and market-share figures change without notice, so verify current documentation before relying on anything here. Sourcing is listed under Sources and References, which separates primary provider documentation from secondary reporting.
Why Index Provenance Changes Your SEO Work
Twenty-five alternative search engines sound like twenty-five places to rank. Several of them query the same underlying provider and present those results through their own interface.
That matters for planning. Work aimed at a provider that licenses its results belongs upstream, at the provider supplying them. Work aimed at an engine running its own crawler needs direct crawlability and, where one exists, its own submission tool.
One caution runs through everything below. Appearing in a provider’s index is not the same as appearing in the products that license from it. Each licensee applies its own ranking, filtering and presentation logic, and AI answer systems add retrieval and synthesis on top. Google states the principle plainly for its own AI features: a page must be indexed and eligible for a snippet, and even then “indexing and serving isn’t guaranteed”.
For the retrieval mechanics behind that, see our guide to how LLMs retrieve, process and cite information from the web, and our overview of what AI search is and how it differs from traditional search.
A Classification Model for Search Providers
Sorting engines into independent and dependent hides the arrangements that matter. Six categories describe the field more accurately.
| Classification | What it means | SEO implication |
|---|---|---|
| Independent web index | The provider crawls and maintains its own searchable web content | Investigate direct crawlability and any submission options |
| Licensed-result search | The provider receives results from another search engine | Optimize for the underlying source, then verify actual visibility |
| Hybrid or blended | The provider combines its own index with licensed results | Track source dependencies and check the output directly |
| Metasearch | Queries other providers at request time | Visibility depends on configuration and the underlying providers |
| AI retrieval and answer system | Retrieves sources then generates or synthesizes a response | Evaluate crawler access, retrieval eligibility and citation behavior separately |
| Platform-specific search | Searches content held inside one platform | Requires a separate platform distribution strategy |
Several engines occupy more than one row, and some move between rows as contracts change. Treat the classifications below as current rather than permanent.
Which Search Engines Maintain Their Own Indexes?
The organizations below operate substantial web-crawling infrastructure. This list is not exhaustive. Smaller and regional indexes exist, the boundaries between independent indexing and licensed results keep shifting, and running a crawler, holding a searchable index and operating a general-purpose search engine are three different things that often get treated as one.
Classification: Independent web index.
Market share: 89.94% worldwide for September 2026 (StatCounter GlobalStats, all platforms, retrieved 16 October 2026).
Webmaster tools: Google Search Console. See our complete guide to every Search Console feature.
SEO note: Google states there are no additional technical requirements for AI Overviews or AI Mode beyond standard indexing and snippet eligibility, and that no new files, markup or structured data are needed to appear in them. For what moves position, see our breakdown of Google ranking factors that matter. For a documented timeline of Google’s ranking changes, including rollout durations and the gaps in the official record, see our guide to Google algorithm updates and their official history.
Google Search Console · Google AI features documentation
Bing
Classification: Independent web index, and the licensor behind several engines below.
Market share: 5.29% worldwide for September 2026 (StatCounter GlobalStats, all platforms).
Webmaster tools: Bing Webmaster Tools, plus IndexNow for submission.
SEO note: Bing visibility can support discovery across several services that use Microsoft search infrastructure or licensed Bing results. It does not guarantee that a page appears at the same position, or appears at all, in another provider’s results or AI-generated answers. Each licensee applies its own logic. Bing Webmaster Tools remains a sensible early priority for most sites, and it appears in our list of free SEO tools professionals use.
Bing Webmaster Tools · IndexNow
Yandex
Classification: Independent web index.
Market share: 1.3% worldwide for September 2026 (StatCounter GlobalStats, all platforms). Country-level reporting places it well ahead of Google inside Russia, with figures varying by measurement source and month.
Webmaster tools: Yandex Webmaster. Accepts IndexNow.
SEO note: Relevant with a Russian-language audience. Russian-language content and Yandex Webmaster verification are separate work from anything done for Google. Hosting location affects latency and user experience, which is distinct from whether Yandex will index a page.
Baidu
Classification: Independent web index.
Market share: 0.44% worldwide for September 2026 (StatCounter GlobalStats, all platforms). Country-level figures for China vary widely between sources, so treat any single number with caution.
Webmaster tools: Baidu Search Resource Platform (Baidu Ziyuan).
SEO note: Three separate questions get conflated here. Whether Baidu can crawl and index your pages is a technical matter of accessibility from mainland networks. Whether you may lawfully operate a commercial site served to mainland users involves ICP licensing and hosting compliance, which is a regulatory question rather than an indexing one. Whether you then rank involves simplified Chinese content and the usual relevance work. Address them in that order, and take legal advice on the second.
Baidu Search Resource Platform
Brave Search
Classification: Independent web index, by Brave’s own account.
Market share: Not separately reported by StatCounter.
Webmaster tools: None public.
Index notes: Brave reports that it removed its last Bing API calls from web results in April 2023, describing those calls as roughly 7% of query results beforehand, and publishes an independence score it says rose from 87% to 93% within a year of launch. Image and video results followed later in 2023. These figures are self-reported and have not been independently audited.
SEO note: No submission tool exists, so Brave finds pages by crawling. Standard crawlability work covers it, and our technical SEO audit checklist covers the ground. Kagi lists Brave among its sources, so Brave visibility can carry into Kagi.
Mojeek
Classification: Independent web index.
Market share: Not separately reported.
Webmaster tools: None, though Mojeek responds to direct contact.
Index notes: Mojeek crawls from the UK with MojeekBot and serves results from its own index. The index is far smaller than Google’s, so long-tail coverage thins out.
SEO note: Useful as a discoverability check rather than a traffic source. A result for your domain demonstrates that you are discoverable in an index with no Google or Microsoft dependency. It does not by itself establish when or how MojeekBot reached you. Server log analysis is what establishes direct crawl activity.
Seznam
Classification: Independent web index, regional.
Market share: Reported between roughly 10% and 15% of Czech search, varying by source and month.
Webmaster tools: Seznam Webmaster.
Index notes: Seznam operates its own crawler, SeznamBot, and its own index for Czech-language search. Most English-language roundups omit it despite a domestic position larger than several engines they include.
SEO note: Worth dedicated work for Czech-market sites and nothing otherwise.
Yep
Classification: Independent web index, in beta.
Market share: Not separately reported by StatCounter. Yep remains in beta, and we found no traffic estimate with a stated methodology and reporting period, so we give no figure.
Webmaster tools: None. Crawled by AhrefsBot.
Index notes: Ahrefs launched Yep in beta in June 2022. Its search API documentation cites a 100 billion page index built from AhrefsBot’s crawl. Index size claims from any provider are self-reported.
SEO note: We found insufficient reliable traffic data to justify prioritizing Yep as an acquisition channel. The reason to care is AhrefsBot, which feeds both this product and the tooling covered in our Ahrefs versus Semrush comparison. Blocking AhrefsBot to conserve crawl budget also restricts Yep inclusion.
Engines Serving Licensed or Blended Results
Each engine below sources some or all web results from another provider. Optimizing for the upstream provider supports discovery here, and then you verify what the engine shows.
DuckDuckGo
Classification: Hybrid. Own crawler and indexes, with traditional links licensed.
Market share: 0.74% worldwide for September 2026 (StatCounter GlobalStats, all platforms).
Index notes: DuckDuckGo’s own help pages state that it maintains “our own crawler (DuckDuckBot) and many indexes to support our results”, while describing traditional links and images as something it largely sources from Bing. Instant Answers draw on further sources including Wikipedia and specialized providers. Requests to partners are proxied through DuckDuckGo servers with no unique identifiers passed.
SEO note: Bing visibility can influence DuckDuckGo web-result visibility, and rankings may differ because DuckDuckGo synthesizes multiple sources and applies its own presentation. Treat Bing work as the main lever and verify by searching DuckDuckGo directly.
DuckDuckGo results sources documentation
Yahoo
Classification: Licensed-result search.
Market share: 1.36% worldwide for September 2026 (StatCounter GlobalStats, all platforms).
SEO note: Yahoo has served Microsoft-supplied web results under a long-running agreement. Bing Webmaster Tools is the practical lever. Confirm placement by searching Yahoo directly rather than assuming parity with Bing.
Startpage
Classification: Licensed-result search, no index of its own.
SEO note: Startpage uses results from external search partners, including Google, and passes queries anonymously. Strong Google visibility supports discoverability on Startpage. Identical rankings, coverage and update timing should not be assumed, since localization, filtering and provider-specific handling can all differ.
Ecosia and Qwant
Classification: Hybrid and actively changing.
Index notes: Both have historically served licensed results, and both are shareholders in European Search Perspective, the joint venture formed in November 2024 to build a European index known as Staan. Reporting describes Qwant as serving part of its results from that index and the remainder from licensed sources, with the index still partial. The current ratio is not publicly documented and changes as the index grows.
SEO note: Do not treat either as a fixed Bing interface. Bing work supports visibility today. Re-check both engines directly if European traffic matters to you, since the mix is in transition.
Swisscows
Classification: Licensed-result search with reported partner blending.
SEO note: Swisscows markets family-safe search from Switzerland and filters adult content. Its result sourcing is described by secondary reviews rather than detailed provider documentation, so verify directly. The content filter excludes adult and adjacent verticals whatever their ranking upstream.
Metasearch Engines
These query other providers when a user searches. Visibility depends on the underlying providers and on each operator’s configuration.
Kagi
Classification: Metasearch with a supplementary own crawler.
SEO note: Kagi charges a subscription, which removes advertising from its ranking incentives. It has listed Brave, Mojeek and Yandex among its sources and operates its own Teclis crawler for small-site discovery. Published source lists change, so check Kagi’s current documentation. Its audience skews technical, which matters more than raw volume for developer and B2B products.
SearXNG
Classification: Metasearch, self-hosted.
SEO note: SearXNG is software rather than a service. Each operator configures which providers to query, so no two instances behave alike and nothing here is directly actionable.
AI Retrieval and Answer Systems
Three questions matter for each system, and they are separate: whether its crawler can reach you, whether you are eligible for retrieval, and whether it cites you. No provider guarantees the third from the first.
Google AI Overviews and AI Mode
Retrieval basis: Google’s standard index. Its documentation states a page must be indexed and eligible to be shown with a snippet, adds that there are no additional technical requirements, and states that “indexing and serving isn’t guaranteed”. Google also says no new machine-readable files, AI text files, markup or special schema.org structured data are needed to appear in these features.
SEO note: One vendor analysis by Evertune, reported by Adapt Worldwide, compared late April with late May 2026 and found unique cited URLs in AI Mode down 59% and Gemini down 33%. That is a single commercial study over roughly one month, with no published methodology detail and no peer review, so treat it as a directional signal rather than an established trend. Our Google AI Mode guide and AI Overviews guide cover these surfaces in depth.
ChatGPT Search
Retrieval basis: OpenAI documents OAI-SearchBot as the crawler used to surface websites in ChatGPT’s search features, and recommends allowing it to help a site appear in search results. Notably, OpenAI states that opted-out sites can still appear as navigational links while being excluded from ChatGPT search answers. OpenAI has not publicly documented the full architecture of ChatGPT Search retrieval, its coverage, or the relative contribution of any external search providers.
SEO note: Allowing OAI-SearchBot supports discovery and potential inclusion. It does not guarantee citation. Blocking it restricts inclusion in search answers rather than removing every possible appearance. OpenAI notes that robots.txt changes take around 24 hours to take effect. Our OAI-SearchBot configuration guide covers crawl access, and how ChatGPT Search works covers retrieval.
Microsoft Copilot
Retrieval basis: Copilot can use Bing-powered web search to ground answers. Copilot ships in several products and contexts, including experiences that draw on user-provided files or connected organizational content, so available sources vary by product.
SEO note: Bing indexing may support visibility in web-grounded Copilot answers. It does not guarantee citation, and it does not define every source a given Copilot experience can use. Most alternative-engine roundups omit Copilot, which is a meaningful gap given its distribution through Windows and Edge.
Perplexity
Retrieval basis: Perplexity operates PerplexityBot and describes maintaining its own index, having started on licensed search in 2022. Reports differ on how much it still calls external APIs, and Perplexity has not published a breakdown.
SEO note: Blocking PerplexityBot may restrict direct crawling and reduce opportunities for inclusion. See our Perplexity citation guide, and our comparison of ChatGPT, Perplexity and AI Overviews for where to spend effort.
Regional Engines Worth Separate Evaluation
Naver, South Korea
Market share: Reported figures differ sharply. StatCounter’s August 2026 data places Naver below Google in Korea, while Korea’s domestic measurer InternetTrend reported Naver well ahead for March 2026. The gap reflects methodology: referral-based measurement undercounts engines that resolve queries inside their own properties, and Naver integrates its own blog, café and shopping services into results.
SEO note: Naver prominently integrates its own content services, so Korean search strategy should evaluate those surfaces separately from your own domain. Publishing into Naver properties is a distinct channel decision rather than a ranking tactic, and the relative weighting Naver applies is not publicly documented.
Chinese Engines Beyond Baidu
Market share: Country-level reporting for China varies substantially by source. StatCounter’s August 2026 China table placed Baidu first, with Bing second and ahead of both Haosou and Sogou. Other published estimates differ by wide margins, and measurement inside China is less reliable than elsewhere.
SEO note: Sogou appears on most alternative-engine roundups while Haosou, which StatCounter ranks higher, appears on almost none. Bing’s reported Chinese share also exceeds common assumptions. For China work, start with Baidu and treat Bing as a secondary surface you may already cover.
Platform-Specific Search, Which Is Not Web Search
Several roundups count platforms that search their own content. The optimization work differs completely, so the distinction is worth keeping.
| Category | Commonly listed examples | Why it needs a different strategy |
|---|---|---|
| Platform-specific search | X, LinkedIn, SlideShare, YouTube, Pinterest | Searches its own database. Visibility comes from publishing on the platform. |
| Archives | Wayback Machine | Retrieves pages already captured rather than crawling for relevance. |
| Computational answer engines | WolframAlpha | Computes from curated datasets and returns no web ranking. |
| Licensed media search | Openverse | Indexes openly licensed media within a narrow scope. |
These services use different discovery and retrieval models from general-purpose web search. Reaching an audience through them may require platform publishing, archive availability, licensed-media inclusion or another service-specific approach rather than conventional website SEO.
Verified Closures
Roundups age, and some entries no longer operate. The table below lists only closures confirmed by multiple independent reports.
| Engine | Status | Date |
|---|---|---|
| Ask.com | IAC closed the search business, ending the Ask Jeeves lineage after nearly 30 years | 1 May 2026 |
| Neeva | Consumer search closed; Snowflake acquired the company and retained the team rather than the engine | 2 June 2023 |
| Gigablast | Stopped serving without announcement after about 21 years running its own crawler and index | April 2023 |
Removed for lack of verification: several engines commonly listed as closed, including Phind, Mullvad Leta, Stract and Arc Search, could not be confirmed against primary announcements or multiple independent reports at the time of writing. Reports also describe You.com shifting away from consumer search toward enterprise products, which we have not confirmed from the company directly. Those entries are omitted rather than repeated. If you rely on any closure claim, check the provider’s own site before citing it.
Market Share, With Provenance
| Engine | Worldwide share | Classification |
|---|---|---|
| 89.94% | Independent index | |
| Bing | 5.29% | Independent index, licensor |
| Yahoo | 1.36% | Licensed results |
| Yandex | 1.30% | Independent index |
| DuckDuckGo | 0.74% | Hybrid |
| Baidu | 0.44% | Independent index |
Data provenance: StatCounter GlobalStats, Search Engine Market Share Worldwide, chart labelled September 2026, default all-platforms view, read directly from the source on 16 October 2026 and confirmed by a second retrieval the same day. StatCounter derives shares from tracked page views across its network rather than from provider disclosures, so its figures differ from panel-based or provider-reported measurement. Device-split and country views return materially different numbers, and Bing’s desktop share runs above its all-platform figure. Re-retrieve before citing, since the chart advances monthly.
StatCounter Search Engine Market Share
How to Verify Search Engine Coverage for Your Website
Every claim in this guide describes general provider behavior. Your site’s actual coverage needs measuring. Four checks, in order.
1. Confirm crawler access, then confirm crawler activity. Your robots.txt states what you permit. Your server logs state what arrived. Filter logs by user agent for Bingbot, OAI-SearchBot, PerplexityBot, MojeekBot and AhrefsBot, and verify the requesting IP resolves to the stated operator, since user agents are trivially spoofed. Permission without log evidence tells you nothing about coverage.
2. Separate indexing from ranking. Use supported site-restricted searches as a preliminary discoverability check rather than as proof of complete indexing, since they return neither a full nor an authoritative inventory. Where a provider offers webmaster tools, use those for indexing diagnostics. Then test representative target queries to evaluate real search visibility. A page can be discoverable everywhere while ranking nowhere.
3. Check the engines directly rather than inferring from upstream. Run the same query on Bing, then on DuckDuckGo, Yahoo and Ecosia. Differences may reflect ranking systems, index coverage, localization, filtering, source blending, query interpretation or update timing. Record the query, location, device and test date before drawing any conclusion about provider relationships.
4. Record referral traffic by provider. Analytics referral data shows which engines and AI surfaces send real sessions. Many AI answer surfaces send little or no referral traffic even while citing you, so pair this with manual citation checks on your priority queries. Our analysis of zero-click searches covers that gap.
Repeat quarterly. Provider arrangements change without announcement, which is the recurring lesson of every item above.
Where to Spend Effort, Conditionally
No action below is universally correct. Each depends on your audience and on current provider coverage you verify yourself.
| Action | What it can support | When it is worth it |
|---|---|---|
| Verify in Bing Webmaster Tools | Discovery across services using Microsoft infrastructure or licensed Bing results | Most sites, as an early low-cost step |
| Submit through IndexNow | Faster notification to Bing and Yandex | Sites publishing or updating frequently |
| Audit robots.txt for AI crawlers | Eligibility for inclusion in ChatGPT search and Perplexity | Any site wanting AI-surface visibility |
| Leave AhrefsBot unblocked | Yep inclusion, plus your own Ahrefs data | Unless crawl budget is a measured problem |
| Check an independent index quarterly | Evidence of discoverability without Google or Microsoft dependency | Sites treating concentration as a risk |
| Add Yandex, Baidu, Naver or Seznam | Regional visibility | Only with an audience in that market |
Frequently Asked Questions
Which search engines maintain independent web indexes?
Google, Microsoft, Yandex, Baidu, Brave, Mojeek, Seznam and Ahrefs are among the organizations operating substantial web-crawling infrastructure, and smaller projects such as Marginalia also crawl independently. This is not a complete list, and the boundary between independent indexing and licensed results keeps shifting as agreements change. Check an engine’s own documentation rather than relying on any roundup.
Does DuckDuckGo use Bing exclusively?
No. DuckDuckGo’s help pages state that it maintains its own crawler, DuckDuckBot, and many indexes supporting its results, while describing traditional links and images as largely sourced from Bing. Instant Answers draw on further sources including Wikipedia. Bing visibility can influence DuckDuckGo web results, and rankings may differ.
Does Bing indexing guarantee Copilot citations?
No. Copilot can use Bing-powered web search to ground answers, so Bing indexing may support visibility in web-grounded responses. Copilot ships across several products with differing available sources, including user-provided and organizational content, and Microsoft documents no citation guarantee from indexing alone.
Can ChatGPT cite a page that OAI-SearchBot has not crawled?
OpenAI states that sites opted out of OAI-SearchBot can still appear as navigational links while being excluded from ChatGPT search answers. OpenAI has not published the full retrieval architecture, so the complete set of circumstances under which a page may surface is not publicly documented. Allowing the crawler is the documented path to inclusion in search answers.
Is allowing an AI crawler enough to appear in AI answers?
No. Allowing a provider’s crawler supports direct discovery and potential inclusion through its documented crawling pathway, and some services also retrieve through licensed results or other sources, so crawler access is one route rather than a universal requirement. Google states that pages must be indexed and snippet-eligible for its AI features and that indexing and serving are not guaranteed. OpenAI recommends allowing OAI-SearchBot to help a site appear without promising placement. Treat crawler access, retrieval eligibility and citation as separate questions.
How do hybrid search providers combine independent and licensed results?
Arrangements vary and few are fully disclosed. DuckDuckGo describes running its own crawler and indexes alongside licensed traditional links. Qwant and Ecosia are building the Staan index through their European Search Perspective joint venture while still serving licensed results, with the proportion undocumented and changing. Because the mix shifts, verify by running the same query across the engines that matter to you rather than assuming a fixed relationship.
Do I need special markup to appear in Google AI Overviews?
No. Google states there are no additional requirements beyond standard indexing and snippet eligibility, and that no new machine-readable files, AI text files, markup or special schema.org structured data are needed for AI Overviews or AI Mode.
How many search engines should I actively work on?
For most English-language sites, Google for volume and Bing as a low-cost second surface that can support discovery across several licensees. Add a regional engine where you have a regional audience. Verify the result in each engine rather than assuming upstream work carries through.
Is Ask.com still running?
No. IAC closed Ask.com on 1 May 2026, ending the Ask Jeeves lineage after nearly three decades. Roundups still listing it predate that closure.
Limits of This Guide
Search providers disclose their retrieval architectures selectively, so parts of this picture rest on provider statements that cannot be independently verified.
Self-reported figures. Brave’s independence percentages and Yep’s index size come from those companies with no external audit.
Undisclosed arrangements. OpenAI has not published the composition of ChatGPT Search retrieval. Perplexity has not published its reliance on external APIs. The Staan-to-licensed ratio at Qwant and Ecosia is not public.
Measurement disagreement. Market share depends on method. StatCounter’s page-view sampling produces different numbers from panel-based or provider-reported measurement, and country-level figures for Korea and China diverge widely between sources.
No ranking guarantees anywhere. No provider in this guide promises that indexing, crawler access or upstream ranking produces placement or citation in its products. Every recommendation here should be validated against your own logs, results and referral data.
Sources and References
Primary provider documentation
- Where DuckDuckGo gets its results (DuckDuckGo Help Pages)
- OpenAI crawler documentation (OAI-SearchBot, GPTBot, ChatGPT-User)
- AI features and your website (Google Search Central)
- Brave Search index independence announcement (Brave)
- Yep Search API documentation (Ahrefs)
- Bing Webmaster Tools (Microsoft)
- IndexNow protocol
- Baidu Search Resource Platform
- Yandex Webmaster
Measurement
Secondary reporting, used only where provider documentation is unavailable
TL;DR Many alternative search engines serve results supplied by larger providers rather than crawling the web themselves. Google, Microsoft, Yandex, Baidu, Brave, Mojeek, Seznam and…
