Baiduspider

TL;DR — Baiduspider is Baidu’s web crawler — the bot that visits your pages, downloads them for Baidu Search. If mainland China is an important market, Baiduspider belongs in your crawl monitoring. Check Baidu’s current webmaster documentation before relying on engine-specific behavior, especially for JavaScript rendering and crawler controls.

What Baiduspider is

Google doesn’t work inside mainland China. Baidu does — it’s the search engine most people there use, and Baiduspider is its crawler. When Baiduspider visits your site, it follows links, reads your sitemap, downloads your pages, and hands them off to build Baidu’s index — the database Baidu pulls from when someone searches. If Baiduspider can’t reach and read a page, that page won’t show up in Baidu.

In Baidu’s own words (from its official English FAQ): “Baiduspider is Baidu search engine program which is used to visit pages on the internet and build information into Baidu index. This enables users to locate your site when they perform a search.” That’s the same job Googlebot does for Google and Bingbot does for Bing.

Why it might matter to you

Baiduspider only really matters if you want to be found inside mainland China. Baidu still leads there while Google barely registers — so a China-facing site (or a global site with a China audience) needs Baiduspider the same way a Western site needs Googlebot. If you don’t serve customers in China at all, Baiduspider is just another bot hitting your server, and some sites choose to block it to save resources.

The catch: Baiduspider is bad at JavaScript

This is the big one. Googlebot runs a modern version of Chrome and reads most JavaScript-heavy sites fine. Baiduspider does not — it’s widely reported to be poor at running JavaScript. If your important content or links only appear after JavaScript runs, Baidu may never see them, even if the same page ranks fine in Google. The safe move for a China-facing site: make sure your real content is in the plain HTML that comes back before any JavaScript runs.

How to help (or stop) Baiduspider

• To let it in: link to your important pages, submit a sitemap in Baidu Search Resource Platform (Baidu’s version of Search Console), and don’t accidentally block those pages in robots.txt.
• To keep it out of parts of your site: use robots.txt — Baidu says it follows the standard rules there.
• To slow it down: Baidu Search Resource Platform has a crawl-frequency area where you can see and cap how much Baiduspider crawls.

Want the technical version — the exact user-agent strings, how to verify a real Baiduspider, the robots.txt quirks, and the crawl-frequency controls? Switch to the Advanced tab.

TL;DR — Baiduspider is Baidu’s crawler — the Chinese-market Googlebot. It runs named variants sharing the Baiduspider token family (general PC/mobile, plus -image, -video, -news, -favo, -cpro, -ads), and a Baiduspider-render JavaScript variant that Baidu documents only in Chinese. Verify it by bidirectional DNS to *.baidu.com/*.baidu.jp — Baidu’s own FAQ warns about impersonators. Baidu claims it “strictly complies with robots.txt protocol,” with a named exception: Baiduspider-cpro and Baiduspider-ads run under commercial agreements and may follow different access rules; it also doesn’t document the non-standard crawl-delay as a reliable control. Because rendering behavior can change, server-rendered HTML is the conservative default for critical content. Verify current controls in Baidu Search Resource Platform (Ziyuan).

What Baiduspider actually is

Baiduspider (百度蜘蛛) is the crawler operated by Baidu, the search engine that dominates inside mainland China, where Google is largely inaccessible behind the Great Firewall. Its job is the standard crawl → index → serve pipeline: discover URLs, fetch pages, and build the index Baidu Search draws from. Baidu’s official English FAQ, “FAQs of Baiduspider” on help.baidu.com, is the rare piece of primary English-language documentation here, and it’s the source for most of the direct quotes below.

The asymmetry that frames everything: Baidu matters domestically, not globally. StatCounter’s June 2026 China snapshot put Baidu at 47.44%, Bing at 23.24%, Haosou at 14.02%, Sogou at 2.62%, and Google at just 2.28% (StatCounter, Search Engine Market Share China). Treat that as a snapshot, not a constant — StatCounter’s China numbers swing meaningfully month to month depending on device mix and methodology (Baidu has been reported anywhere from roughly 40% to 65% across different 2025–2026 snapshots), so re-check the live figure rather than quoting one month forever. The takeaway holds regardless: Baidu leads inside China while Google leads the rest of the world, which is why Baiduspider optimization is treated as its own international-SEO specialty rather than a Googlebot afterthought. Our market-specific SEO hub covers where Baidu fits alongside Yandex, Naver, and the other regional engines.

User-agent strings and crawler variants

Baidu’s English FAQ names the crawler variants directly. Reproduced verbatim (and yes, Baidu’s own copy repeats “Baiduspider” for both PC and mobile rather than giving distinct strings):

Name of Products | User-agent PC search | Baiduspider Mobile search | Baiduspider Image search | Baiduspider-image Video search | Baiduspider-video News search | Baiduspider-news Baidu bookmark | Baiduspider-favo Union baidu | Baiduspider-cpro Business search | Baiduspider-ads other search | Baiduspider

So the officially named variants in the English FAQ are Baiduspider-image, -video, -news, -favo (bookmark), -cpro (the “union”/ad-network crawler), and -ads (business search) — all sharing the Baiduspider family token. The literal general PC user-agent string, given on Baidu’s own site, is:

Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)

The mobile UA embeds Baiduspider/2.0 inside an Android WebKit string. Every declared variant points to the same http://www.baidu.com/search/spider.html reference URL — the Baidu equivalent of Googlebot’s google.com/bot.html.

Baiduspider-render — the JS variant Baidu only documents in Chinese

Here’s a subtlety worth getting right, because a lot of write-ups get it wrong: the English FAQ table above does not list a “Baiduspider-render” variant, which has led some guides to call it merely third-party folklore. It isn’t. Baidu documents Baiduspider-render officially — just on its Chinese-language Ziyuan Academy page, “【官方说法】只需两步，正确识别百度蜘蛛（User-Agent）” (“[Official statement] Two steps to correctly identify Baiduspider”). That page splits UAs into three channels — 移动 (mobile), PC, and 小程序 (mini-program) — and lists an alternate render UA alongside the plain one:

Mozilla/5.0 (compatible; Baiduspider-render/2.0; +http://www.baidu.com/search/spider.html)

with a matching mobile string and a mini-program variant (Baiduspider-render/2.0;Smartapp). So Baiduspider-render is officially documented — just Chinese-only, absent from the English FAQ. What Baidu does not publish is a spec for what “-render” actually does differently; unlike Google, Baidu doesn’t document its JavaScript rendering capabilities. That gap is where the JS-rendering caveat below comes from.

How to verify Baiduspider (and not a spoofed bot)

The user-agent string is trivially spoofed — anyone can send a request that says Baiduspider. Baidu’s own FAQ warns about exactly this, describing “spammers or other trouble makers who pretend to be Baiduspider,” and its guidance is to verify with DNS, not the string. From the FAQ, verbatim:

We recommend using reverse DNS lookup to verify Baiduspider. Verification methods are different under linux/windows/os environments.

It gives per-OS commands — host on Linux, nslookup on Windows, dig -x on macOS — with worked examples, and the rule that matters:

The hostname of Baiduspider is *.baidu.com or *.baidu.jp. Others are fake hostnames.

The two *.baidu.com and *.baidu.jp domains are the quirk to remember — Baidu crawls from both a .com and a .jp reverse-DNS namespace, so a legitimate Baiduspider hostname can end in either. (One of Baidu’s own worked examples even resolves to BaiduMobaider-119-63-195-254.crawl.baidu.jp — note “BaiduMobaider,” not “Baiduspider,” in that hostname; that’s a real quirk in Baidu’s documentation, not a typo to silently correct.)

Baidu’s English FAQ spells out only the reverse-lookup half in its prose, but its Chinese Ziyuan Academy page is more explicit: it names the whole process “双向DNS解析认证” (bidirectional DNS resolution authentication) and breaks it into “第一步：DNS反查IP” (Step 1: reverse-lookup the IP) and “第二步：对域名运行正向DNS查找” (Step 2: run a forward DNS lookup on the hostname). That’s the complete, correct method — reverse-lookup the IP to confirm a *.baidu.com/*.baidu.jp hostname, then forward-lookup that hostname and confirm it resolves back to the same IP. It’s the same two-step approach you’d use to verify Googlebot or Bingbot — the commands are in the Scripts tab.

Baiduspider and robots.txt

Baidu makes an unusually strong, quotable compliance claim in its own FAQ:

Baidu strictly complies with robots.txt protocol.

And it supports the standard controls you’d expect. Per Baidu’s Ziyuan Academy material, Baiduspider supports wildcard matching with * and $ in robots.txt paths (parity with Google and Bing), and does exact, case-sensitive path matching. You can set per-user-agent rules — Baidu’s FAQ gives worked examples, including this one that blocks everything for general Baiduspider but allows the image crawler into /image/:

User-agent: Baiduspider
Disallow: /
User-agent: Baiduspider-image
Allow: /image/

(Baidu also notes that Baiduspider-video doesn’t currently support these rules.)

The named exception to “strict compliance”

The interesting tension: Baidu claims strict compliance but documents a named, official exception. Baiduspider-cpro (the ad-network crawler) and Baiduspider-ads (business search) operate under separate commercial agreements and ignore robots.txt by design. Verbatim from the FAQ: “Baiduspider-cpro will not work on the records set by robots.txt” and the same for -ads. So the honest framing isn’t the vague “Baidu doesn’t always respect robots.txt” you’ll see on aggregator sites — it’s that Baidu’s standard crawlers comply, with two contractually exempt ad crawlers as the one concrete, documented carve-out.

A separate, narrower point that’s easy to conflate: Baiduspider does not honor the non-standard crawl-delay directive. Search Engine Journal’s Abby Hamilton documented this in The Modern Guide To Robots.txt (Nov 2024) — her crawler comparison table lists “Baidu | Baiduspider | No” for crawl-delay support (Yandex’s YandexBot, by contrast, is a “Yes”). That’s a different claim from standard Allow/Disallow compliance, and worth keeping straight so you don’t accidentally contradict Baidu’s own robots.txt claim.

A real example: baidu.com’s own robots.txt

Baidu practices robots.txt segmentation on its own domain. Its live robots.txt gives its own Baiduspider a shorter disallow list than it gives Googlebot, MSNBot, Sogou, and Youdao (which additionally get /shifen/, /homepage/, and /cpro blocked), and blocks everything (Disallow: /) for any unnamed user-agent by default:

User-agent: Baiduspider
Disallow: /baidu
Disallow: /s?
Disallow: /ulink?
Disallow: /link?
Disallow: /home/news/data/
Disallow: /bh

One operational detail from the FAQ worth knowing when you’re editing rules: if you tighten robots.txt after Baidu has already indexed a site, “it usually takes 48 hours for the updated robots.txt to take effect.”

Does Baiduspider render JavaScript?

This is Baiduspider’s biggest practical limitation relative to Googlebot, and it’s worth being precise about what’s official and what isn’t. Baidu’s English FAQ doesn’t address JavaScript rendering at all — there’s no capability spec the way Google publishes one for Googlebot. What we have instead is consistent industry testing. Dan Taylor put it bluntly in Search Engine Journal’s Google vs Baidu comparison (Mar 2021): “Baidu is notoriously bad when it comes to crawling JavaScript, so make sure that all your important content and links are served in plain HTML on both the mobile and desktop versions of your website.”

That aligns with the position this site already takes. Our contentful SEO coverage notes that Bing, Yandex, and Baidu “may not index [client-side-rendered JS] at all,” and that dynamic rendering is deprecated as a general Google technique. Put the three big crawlers on a spectrum: Googlebot does full evergreen-Chromium rendering, Bingbot does solid Chromium/Edge rendering, and Baiduspider is the weakest of the three at JavaScript — with plain, server-rendered HTML being the universally safe fallback across all of them.

Practical guidance for a China-facing site: default to server-side rendering or static HTML so critical content, links, and metadata exist in the initial response before any JavaScript runs. Treat client-side-only content as an indexing risk for Baidu specifically, independent of whether it ranks fine in Google. Frame SSR as an industry-consensus best practice here — not as an official Baidu recommendation, because no primary Baidu statement on JS rendering exists to cite.

Controlling Baiduspider’s crawl frequency

Baidu presents crawl frequency as algorithmically self-adjusted, not something you set directly. From the FAQ, verbatim:

In order to ensure the search results cover most of your pages, Baiduspider must keep the crawling at a certain level. We have been trying our best to avoid increasing the loading to your servers, and to adjust the frequency based on combined factors, such as your server’s capability, your site’s quality and the update frequency of your site.

That’s the same server-capability-plus-demand logic Google and Bing describe. The FAQ names a feedback form (webmaster.baidu.com/feedback/index) as the official lever for reporting over-crawling — “If you find any unreasonable access from Baiduspider, please inform us.”

At the product level, Baidu Search Resource Platform (百度搜索资源平台, at ziyuan.baidu.com — the console historically called “Baidu Webmaster Tools”) is Baidu’s equivalent of Google Search Console and Bing Webmaster Tools. Its manual index confirms a “抓取频次” (Crawl Frequency) product area where site owners can view and cap Baiduspider’s crawl rate directly, alongside robots and crawl-diagnostics tools. Treat the feedback form as the official escalation path in Baidu’s English docs and the Ziyuan Crawl Frequency dashboard as the practical, UI-level control. (The exact in-product menu path sits behind a logged-in Ziyuan account, so confirm the current UI there before quoting specific steps.)

hreflang and other Google-standard signals

Signals that are standard for Google don’t all carry over to Baidu — and this is an industry/practitioner finding, not something either company documents. Google’s own localized-versions documentation is scoped entirely to Google Search and makes no claim about whether Baidu, Bing, or Yandex honor hreflang. On this site, our hreflang guide states it plainly: Baidu doesn’t support hreflang at all — the fallback is to set an accurate content-language and <html lang> so Baidu can at least read the page’s language. That’s additive to my own Google-scoped hreflang work (the Ahrefs hreflang tags guide and the study of 374,756 domains, both of which discuss Google’s implementation only) rather than a contradiction of it.

No English rep culture — why the FAQ is unusually valuable

One thing that genuinely sets the Baidu ecosystem apart: unlike Google (Gary Illyes, John Mueller), Bing (Fabrice Canel), or Yandex, Baidu maintains no visible English-language public-rep Q&A culture. There’s no named spokesperson doing SEO community back-and-forth. Baidu’s clearest official English voice is the unsigned, institutional help.baidu.com FAQ. That’s why that FAQ — and the Chinese-language Ziyuan Academy pages behind it — are so valuable: most operational knowledge about Baiduspider in the English-speaking SEO world comes from practitioner testing (server-log analysis, DNS-verification exercises, robots.txt trial and error) rather than official rep commentary. Michael Bonfils summarized the broader lesson in Search Engine Land’s Baidu vs. Google: “SEO strategies that work for Google may not always translate directly to success on Baidu, China’s dominant search engine.”

For where Baiduspider sits in the bigger picture — the crawler family, Googlebot and Bingbot as peers, the user-agent concept, and robots.txt — the crawling hub ties it together, and international SEO is the home for the China-market context.

AI summary

A condensed take on the Advanced version:

• Baiduspider = Baidu’s web crawler — the Chinese-market Googlebot. Baidu dominates inside mainland China (StatCounter June 2026: Baidu ~47%, Google ~2%), so it’s an international-SEO specialty, not a Googlebot afterthought. That share swings month to month — treat it as a snapshot.
• Variants share the Baiduspider token family: general PC/mobile plus -image, -video, -news, -favo (bookmark), -cpro (ad network), -ads (business). General PC UA: Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html).
• Baiduspider-render is officially documented — but only on Baidu’s Chinese-language Ziyuan Academy page, absent from the English FAQ. What “-render” does differently is not spec’d.
• Verify by bidirectional DNS to *.baidu.com or *.baidu.jp (Baidu crawls from both): reverse-lookup the IP, then forward-lookup the hostname back to the same IP. The user-agent is spoofable; Baidu warns about impersonators.
• robots.txt: Baidu claims it “strictly complies,” supports */$ wildcards and case-sensitive matching — with a named exception: Baiduspider-cpro and Baiduspider-ads run under commercial agreements and ignore robots.txt. Separately, Baiduspider does not honor the non-standard crawl-delay (per SEJ’s Abby Hamilton). Rule changes take ~48 hours to take effect.
• JavaScript is the biggest gap vs. Googlebot — no official capability spec; industry consensus (Dan Taylor: “notoriously bad… crawling JavaScript”) is that server-rendered/static HTML is the safe default for China-facing sites.
• Crawl frequency is algorithmic (server capability + site quality + update frequency); report over-crawling via Baidu’s feedback form, and view/cap the rate in Baidu Search Resource Platform (Ziyuan) — Baidu’s Search Console equivalent.
• hreflang isn’t supported by Baidu (industry finding; neither company documents it either way) — fall back to content-language/<html lang>.
• No English rep culture: Baidu’s clearest official English voice is an unsigned FAQ, so practitioner testing carries most operational knowledge.

Official documentation

Primary-source documentation. Baidu’s is split across an English FAQ and Chinese-language Ziyuan (Baidu Search Resource Platform) pages; Google links are for the hreflang-intersection contrast.

Baidu

• FAQs of Baiduspider (help.baidu.com) — Baidu’s English-language official FAQ: what Baiduspider is, the user-agent table, crawl-frequency policy, DNS verification with worked examples, and robots.txt handling. The primary English source.
• 【官方说法】只需两步，正确识别百度蜘蛛（User-Agent）(ziyuan.baidu.com) — Baidu’s Chinese-language official article on identifying Baiduspider: the three UA channels (mobile/PC/mini-program), the Baiduspider-render UAs, and the “bidirectional DNS authentication” method spelled out in full.
• Baidu Search Resource Platform manual index (ziyuan.baidu.com) — the Ziyuan product manual index confirming the 抓取频次 (Crawl Frequency), robots, and 抓取诊断 (Crawl Diagnostics) tool areas.
• baidu.com/robots.txt — Baidu’s own live robots.txt, a real example of per-crawler segmentation.

Google (for the hreflang-intersection contrast)

• Tell Google about localized versions of your page — Google’s hreflang docs, scoped entirely to Google Search; notably says nothing about whether Baidu, Bing, or Yandex honor hreflang.

Quotes from the source

Baidu keeps no named English-language spokesperson, so the “source” here is Baidu’s own unsigned FAQ, plus named industry commentary that stands in for the rep quotes you’d get from Google or Bing. Each link is a deep link that jumps to the quoted passage.

Baidu — “FAQs of Baiduspider” (help.baidu.com, English)

• “Baiduspider is Baidu search engine program which is used to visit pages on the internet and build information into Baidu index. This enables users to locate your site when they perform a search.” Jump to quote
• “In order to ensure the search results cover most of your pages, Baiduspider must keep the crawling at a certain level… to adjust the frequency based on combined factors, such as your server’s capability, your site’s quality and the update frequency of your site.” Jump to quote
• “We recommend using reverse DNS lookup to verify Baiduspider.” Jump to quote
• *“The hostname of Baiduspider is *.baidu.com or .baidu.jp. Others are fake hostnames.” Jump to quote
• “Baidu strictly complies with robots.txt protocol.” Jump to quote

Baidu — Ziyuan Academy (ziyuan.baidu.com, Chinese)

• Original: “二、双向DNS解析认证” — English paraphrase: “Step two: bidirectional DNS resolution authentication,” Baidu’s own name for verifying Baiduspider via a reverse lookup followed by a forward lookup back to the same IP. View the page

Dan Taylor, Search Engine Journal — on JavaScript

• “Baidu is notoriously bad when it comes to crawling JavaScript, so make sure that all your important content and links are served in plain HTML on both the mobile and desktop versions of your website.” (Mar 2021.) Jump to quote

Michael Bonfils, Search Engine Land — on strategy transfer

• “SEO strategies that work for Google may not always translate directly to success on Baidu, China’s dominant search engine.” Jump to quote

Abby Hamilton, Search Engine Journal — on crawl-delay

• Her crawler comparison table lists Baiduspider as not supporting the non-standard crawl-delay directive (“Baidu | Baiduspider | No”). (Nov 2024.) Read the coverage

Should I let Baiduspider crawl my site — and how do I handle it?

Baiduspider is only worth the effort if you actually want China-market visibility. Click through it.

Verify a bot is really Baiduspider

Baidu’s own FAQ recommends reverse DNS to verify Baiduspider, because the user-agent string is trivially spoofed. Do the full bidirectional check Baidu names on its Chinese Ziyuan page: reverse-lookup the IP (the hostname must end in *.baidu.com or *.baidu.jp), then forward-lookup that hostname back to the same IP.

macOS / Linux

# 1) Reverse-DNS the IP from your logs — it must end in baidu.com or baidu.jp
host 123.125.66.120
# → 120.66.125.123.in-addr.arpa domain name pointer Baiduspider-123-125-66-120.crawl.baidu.com

# 2) Forward-DNS that hostname back — it must resolve to the same IP
host Baiduspider-123-125-66-120.crawl.baidu.com
# → Baiduspider-123-125-66-120.crawl.baidu.com has address 123.125.66.120

macOS dig form (as Baidu’s FAQ gives it):

dig -x 123.125.66.120

Windows

nslookup 123.125.66.120
nslookup Baiduspider-123-125-66-120.crawl.baidu.com

If the reverse lookup doesn’t end in baidu.com or baidu.jp, or the forward lookup doesn’t return the original IP, it isn’t Baiduspider — drop or rate-limit it. (Same two-step approach as verifying Googlebot and Bingbot; see those articles’ Scripts tabs for the Google/Bing versions.)

Pull only Baiduspider hits from your server logs

Grab candidate Baiduspider lines from an access log, then verify each IP with the DNS check above — don’t trust the string alone.

# Extract lines whose user-agent claims to be Baiduspider, print unique IPs
grep -i 'baiduspider' /var/log/nginx/access.log \
  | awk '{print $1}' | sort -u

Match a hostname against Baidu’s allowed domains (regex)

Once you’ve reverse-resolved a hostname, this regex confirms it’s in one of Baidu’s two legitimate crawler namespaces (.baidu.com or .baidu.jp) rather than a look-alike:

# Passes only for hostnames ending in .baidu.com or .baidu.jp
echo "Baiduspider-123-125-66-120.crawl.baidu.com" \
  | grep -Eiq '\.baidu\.(com|jp)$' && echo "legit namespace" || echo "FAKE"

Beware look-alikes like baidu.com.evil.example — anchoring the pattern to the end of the string ($) is what stops those from passing.

A robots.txt starting point for Baiduspider

# Let Baiduspider crawl everything except low-value spaces
User-agent: Baiduspider
Disallow: /search
Disallow: /*?*sort=

# Block Baiduspider entirely (China not a target market)
# User-agent: Baiduspider
# Disallow: /

Remember: Disallow controls crawling, not indexing, and rule changes take ~48 hours to take effect. The ad crawlers Baiduspider-cpro and Baiduspider-ads ignore robots.txt by design.

Baiduspider — cheat sheet

User-agent strings and crawler variants

Every declared variant’s UA points to the same reference URL, http://www.baidu.com/search/spider.html — the Baiduspider equivalent of Googlebot’s google.com/bot.html.

robots.txt compliance at a glance

Fast facts

• Verify with bidirectional DNS: reverse-lookup the crawling IP to confirm a hostname ending in *.baidu.com or *.baidu.jp, then forward-lookup that hostname back to the same IP. The user-agent string alone proves nothing.
• robots.txt changes take about 48 hours to take effect on an already-indexed site.
• Baidu supports * and $ wildcard matching in robots.txt paths, and does exact, case-sensitive path matching.
• Crawl frequency is algorithmically self-adjusted (server capability, site quality, update frequency) — view and cap it in Baidu Search Resource Platform (Ziyuan).
• Baiduspider’s biggest practical gap vs. Googlebot is weak JavaScript rendering — server-rendered or static HTML is the safe default for China-facing content.
• Baidu doesn’t support hreflang — the fallback signal is an accurate content-language and <html lang>.

Baiduspider readiness checklist

A quick pass to confirm you’re working with Baiduspider, not accidentally against it — the same ground covered in the Advanced tab, in checklist form.

• robots.txt doesn’t block Baiduspider (or the variant you care about) from anything you want found in Baidu search results.
• A sitemap is submitted in Baidu Search Resource Platform (Ziyuan).
• Bot verification uses bidirectional DNS — reverse lookup to *.baidu.com/*.baidu.jp, then a forward lookup confirming the same IP — never the user-agent string alone.
• Aware that Baiduspider-cpro and Baiduspider-ads ignore robots.txt by design, so blocking those two there won’t actually stop them.
• Any recent robots.txt tightening has had roughly 48 hours to propagate before you re-check indexing status.
• Critical content, internal links, and metadata all exist in the server-rendered HTML response — not injected only by client-side JavaScript.
• content-language and <html lang> are set correctly, since Baidu doesn’t support hreflang.
• Crawl load reviewed in Ziyuan’s Crawl Frequency (抓取频次) area if Baiduspider traffic is straining the server.
• Unreasonable or excessive crawling reported through Baidu’s feedback form (webmaster.baidu.com/feedback/index) rather than guessed at or blocked outright.
• No lingering reliance on Baidu MIP — the framework is retired (Cache phased out June 2020, full service discontinued by late September 2022).

Tools for working with Baiduspider

• DNS Checker — my own tool for the bidirectional DNS check Baidu’s own FAQ recommends. Enter the crawling IP with record type PTR to run the reverse lookup (a genuine Baiduspider hostname ends in .baidu.com or .baidu.jp), then enter that hostname with record type A to confirm the forward lookup resolves back to the same IP. It queries Cloudflare, Google, and Quad9 side by side so you can spot a resolver disagreeing.
• robots.txt Tester — my own tool for checking whether your robots.txt rules actually block or allow Baiduspider the way you intend. Paste your file, enter Baiduspider (or Baiduspider-image, Baiduspider-news, etc.) in the custom user-agent field, and test it against real paths — it shows the exact rule that wins per URL. Remember that Baiduspider-cpro and Baiduspider-ads ignore robots.txt regardless of what the tester shows, since they run under separate commercial agreements.

Baiduspider myths and mistakes to avoid

Myth: “Any user-agent containing ‘Baiduspider’ is really Baidu.” Why it’s wrong: the user-agent string is trivially spoofed, and Baidu’s own FAQ warns about “spammers or other trouble makers who pretend to be Baiduspider.” Do instead: run the bidirectional DNS check — reverse-lookup to a *.baidu.com/*.baidu.jp hostname, then forward-lookup back to the same IP — before you trust (or block based on) any “Baiduspider” hit.

Myth: “Baidu never follows robots.txt, or is generally unreliable about it.” Why it’s wrong: Baidu’s FAQ states “Baidu strictly complies with robots.txt protocol” for its standard crawlers. The real exception is narrow and named — Baiduspider-cpro and Baiduspider-ads run under commercial agreements and ignore robots.txt — not a vague “Baidu ignores robots.” Do instead: cite the specific cpro/ads carve-out, and don’t conflate it with the separate fact that Baiduspider doesn’t honor the non-standard crawl-delay directive.

Myth: “Baiduspider renders JavaScript about as well as modern Googlebot.” Why it’s wrong: industry consensus (Dan Taylor called it “notoriously bad… crawling JavaScript”) is that Baiduspider lags well behind Googlebot’s evergreen Chromium; client-side-only content is a real indexing risk for Baidu even when it ranks fine in Google. Do instead: serve important content, links, and metadata in server-rendered or static HTML for China-facing sites — the safe default across Googlebot, Bingbot, and Baiduspider alike.

Myth: “hreflang works the same for Baidu as it does for Google.” Why it’s wrong: Baidu doesn’t support hreflang at all (an industry finding — Google’s own docs don’t comment on other engines either way). Do instead: fall back to accurate content-language and <html lang> so Baidu can at least read the page’s language; don’t rely on hreflang to steer Baidu.

Myth: “‘Baiduspider-render’ isn’t an officially documented variant.” Why it’s wrong: Baidu’s English FAQ table doesn’t list it, but Baidu’s Chinese-language Ziyuan Academy page documents Baiduspider-render/2.0 as an alternate PC/mobile UA and a mini-program variant. It’s official — just Chinese-only. Do instead: describe it as officially documented (Chinese-only), while noting Baidu publishes no spec for what “-render” actually does differently.

Myth: “Baidu MIP is still a live, recommended ranking lever.” Why it’s wrong: Baidu’s AMP-like MIP (Mobile Instant Pages) framework is legacy. Per Jademond’s MIP glossary, the MIP Cache was phased out in June 2020 and the full MIP service was discontinued by late September 2022 — yet some current-dated agency content still promotes it as live, which is itself a source of confusion. Do instead: treat MIP as retired, and be skeptical of any recent guide that presents it as an active speed/ranking lever.

Test yourself: Baiduspider

Five quick questions on Baidu’s crawler. Pick an answer for each, then check.

Resources worth your time

My related writing

• Hreflang Tags: A Simple (But Complete) Guide — my Ahrefs guide (with Joshua Hardwick) on hreflang; Google-scoped, and the context for why Baidu’s lack of hreflang support is a separate concern.
• Over 67% of Domains Using Hreflang Have Issues — my team’s study of 374,756 domains; Google-implementation focused, useful as the contrast to Baidu’s non-support.
• The Beginner’s Guide to Technical SEO — where crawlers and international signals fit in the bigger picture.

My speaking

• International SEO: The Weird Technical Parts (Pubcon Vegas 2019, SlideShare) — my talk on the technical edge cases of international SEO, the terrain Baidu-market work lives in.

From around the industry

• FAQs of Baiduspider (Baidu, help.baidu.com) — Baidu’s own English FAQ: UA table, DNS verification, robots.txt handling, and crawl-frequency policy. The primary English source.
• Google vs Baidu: Key Differences in SEO Strategy (Dan Taylor, Search Engine Journal) — the “notoriously bad at crawling JavaScript” assessment and the plain-HTML recommendation.
• Baidu vs. Google: Navigating the SEO landscape in 2024 (Michael Bonfils, Search Engine Land) — agency-founder view on why Google tactics don’t transfer directly to Baidu.
• The Modern Guide To Robots.txt (Abby Hamilton, Search Engine Journal) — the crawler comparison table showing Baiduspider doesn’t honor crawl-delay.
• A guide to web crawlers: What you need to know (James Allen, Search Engine Land) — Baiduspider’s place in a broader crawler roundup.
• Understanding Baidu Spiders, User Agents & robots.txt (Jademond Digital) — practitioner deep dive on the Baidu crawler variants and robots handling.
• Baidu MIP (glossary) (Jademond Digital) — the source for the MIP deprecation dates (Cache phased out June 2020; full service discontinued late September 2022).

Baiduspider

Baiduspider is the web crawler operated by Baidu, China's dominant search engine — the Chinese-market equivalent of Googlebot. It discovers, crawls, and indexes pages for Baidu Search, and you manage its access via robots.txt and Baidu's own console, Baidu Search Resource Platform (Ziyuan).

Related: Googlebot, Bingbot, Crawler

Baiduspider

Baiduspider (also written BaiduSpider) is the web crawler run by Baidu (百度), the search engine that dominates inside mainland China, where Google is largely inaccessible. It plays the same role Googlebot plays for Google: it discovers URLs, downloads pages, and builds the index Baidu Search pulls from. Site owners manage it through robots.txt and through Baidu’s own webmaster console — Baidu Search Resource Platform (百度搜索资源平台, at ziyuan.baidu.com), historically called “Baidu Webmaster Tools.”

Baidu runs several named crawler variants that share the Baiduspider token family — general web (PC and mobile), plus Baiduspider-image, -video, -news, -favo (bookmark), -cpro (ad network), and -ads (business search). A JavaScript-rendering variant, Baiduspider-render, is officially documented too — but only on Baidu’s Chinese-language help pages, not in its English FAQ.

Two things distinguish Baiduspider from Googlebot and Bingbot. First, its JavaScript rendering lags well behind: industry consensus is that heavy client-side content is a real indexing risk for Baidu specifically, so server-rendered HTML is the safe default for China-facing sites. Second, Baidu keeps no public English-language rep dialogue the way Google, Bing, and Yandex do — its clearest official English voice is an unsigned FAQ. As with the other engines, verify a bot claiming to be Baiduspider with bidirectional DNS (reverse lookup to *.baidu.com or *.baidu.jp, then a forward lookup back to the same IP) — the user-agent string alone is trivially spoofed.

Related: Googlebot, Bingbot, Crawler