Does Baidu read JavaScript?

Most guides settle this in one line, either “Baidu can’t” or “Baidu struggles.” Few show what that line rests on, and the evidence is older and thinner than they suggest.

This page sets it out by date, then shows what to change on your site. It’s part of our guide to Baidu SEO and how it differs from Google.

Does Baidu run JavaScript at all?

Baidu has a crawler built for it, but we found no statement of what that crawler actually does. We also found no one outside Baidu who can show how far it goes.

On March 24, 2017, Baidu announced a new agent, Baiduspider-render, to access the CSS, JavaScript and images of websites.¹ It started as a beta on a selection of quality sites.¹

The notice’s title calls this rendering crawl, but its text only says Baidu needs to access those files. It doesn’t say which scripts run, how long they get, or which sites are covered.

We found no later Baidu notice that widens the beta, ends it or sets out limits. One bot directory says it was designed to render JavaScript-heavy pages, but cites no test.³¹

What the evidence says, by date

The tests point one way, and the newest Baidu text we found on this is from 2019. Here is what we found, with the weak items labeled.

Date	Source	Kind	What it found
Oct 2016	freeCodeCamp 12	site: check, one client-rendered site	A follow-up test counted 1 Baidu result for preactjs.com, against 17 on Google
Mar 2017	Baidu 1	Official notice	Render agent in beta on selected sites
Aug 2017	Baidu engineer 4	Official talk	Don’t load titles, dates or body text with JS
About Aug 2017	Bartosz Goralewicz 13	Controlled test	No Baidu result: the test site wasn’t indexed there when he published
Nov 2017	Baidu white paper 5	Official guide	JS-loaded mobile pages are more prone to thin content
May 2018	V2EX forum thread 34	Opinions, no logs	Posters disagree: one says Baidu had begun running JS, others say Vue and React pages won’t get indexed
Summer 2018	Unless 9	Controlled test, search excerpt	Baidu didn’t seem to recognize any injected content
Mar 2019	Baidu 23	Official guide	JS content can’t be searched, and content Baidu can’t parse gets discarded
Aug 2019	Chinafy 10	Ranking correlation	No link between dynamic content and Baidu rank
Nov 2019	Botify 11	Vendor claim, no source given	Baidu renders JavaScript, “albeit selectively”
Feb 2020	Salt.agency 8	Controlled test, search excerpt	No for Baidu on all 4 injection methods
2026	Marketing to China, SEO Sherpa 1416	Agency claims, no test	Baidu “struggles” with JS sites

Read it with care, for three reasons:

• Nothing recent. The latest test we found is from February 2020. Baidu may have changed since then, and we found no published retest.
• Test pages. The injection tests use purpose-built pages, which Baidu may treat differently from your site.
• Partial reading. We could read the Salt.agency and Unless results only as search excerpts, so we can’t report their full setup.

Access is one practical barrier. Goralewicz wrote that he couldn’t get his test site indexed on Baidu, and said you need a mainland China phone number to do it.¹³

Our page on the Baidu account and phone number problem covers that.

What Baidu’s own documents tell you to do

Baidu’s written advice is consistent: put what matters in the HTML, and keep JavaScript for interaction. These texts are old, but we found nothing that replaces them.

Part of the page	What Baidu says	Date
Title, date, body text	Don’t load them with JS, it works against the Spider’s assessment 4	2017
Important content and links	Use text, not JS 2	2019
Pages linked only from JS	Baidu may not be able to index them 2	2019
Navigation and body text	Keep them out of Ajax, which is for user interaction 2	2019
Prices	Prices output by JS are unfriendly to the spider and hard to use in search 6	2017
Redirects	A JS redirect Baidu can’t recognize is listed as a crawl anomaly 318	2019

There’s a tension here. The guide, last updated in 2019, says search engines can’t identify content in JavaScript,² two years after the render agent.¹ We found no text from Baidu that reconciles the two.

The advice still holds either way, because HTML works whatever the render agent does.

Does JavaScript hurt rankings on its own?

Nothing we found shows that it does. The risk is content Baidu can’t see, not the presence of scripts.

Chinafy’s 2019 study found no correlation between how dynamic a page was and its Baidu rank.¹⁰

Treat that as a rough signal. It compared rank with page content before and after Ajax, so it didn’t test what Baidu reads. It also dropped 23 all-JavaScript pages, and its write-up still contains unfilled placeholders.¹⁰

Baidu itself tells developers how to use JavaScript well. Its 2019 performance guide recommends defer first, then async, and splitting out the small amount of code the first screen needs.²¹

The danger is a page that looks empty:

• Baidu’s 2023 list of low-quality pages includes pages that open normally but have no content.²⁸
• Its 2017 white paper warns that JS-loaded mobile pages are more prone to thin content.⁵
• Its 2019 guide says very slow pages may also be treated as thin.³

That last point matters for sites hosted abroad. See how to speed up your website in China, and check that no script depends on blocked Google services.

Free check

Is your website visible on Baidu?

Enter your website. We check Baidu, access from China and how your site is set up, then email you the report.

Your websiteCheck my site

Which rendering setup to use for Baidu

Choose any setup that puts your text and links in the first HTML response. The framework doesn’t matter, as long as the output does that.

Setup	Text in first HTML?	For Baidu
Static generation	Yes	Simplest fit for marketing sites and blogs
Server-side rendering (Next.js, Nuxt 14)	Yes	Fits Baidu’s advice, and suits pages that change often
Classic CMS with some scripts	Usually	Fine, but check tabs, “load more” and filters
Prerendering for crawlers only	For crawlers	A workaround, see below
Client-side only	No	The case Baidu’s guides warn against

Prerendering for crawlers. This means detecting bots and serving them a separate rendered copy. Google calls it a workaround, and recommends server rendering, static rendering or hydration.²⁰

We found no Baidu rule that bans or approves it. The 2017 white paper does address pages optimized for crawlers: stay within 128k and put the main content first.⁵

A guest article Baidu published in 2015 favors the same page for people and spiders.¹⁷ If you use prerendering anyway, keep the content identical.

1. Is the text in the source code?

   Yes: check the page weight.

2. No, brochure site or blog

   Static generation

3. No, content that changes often

   Server rendering

4. Rebuild impossible in the short term

   Prerendering for bots, identical content

The trap in server rendering: page size

Server rendering can create a Baidu problem of its own. Baidu’s 2017 white paper says source over 128k may be cut off during crawling, and it advises staying within that.⁵

Its example is telling. A site generated its main content with JS, so it built a crawler version with base64 images inside the HTML, placed before the text. The page reached 164k.

The crawler cut the page off, it was judged thin, and it wasn’t indexed.⁵

Rendered pages often carry inline data and styles as well, so check the size of the source, not just the transfer.

Links and URLs

Render navigation as real <a href> links in the HTML, because pages reachable only through JavaScript may not get indexed.²

The 2016 count of preactjs.com fits that. The person who ran it suspected that engines other than Google indexed the page but didn’t follow its links.¹²

Use real paths and not # routes. The 2015 guest article argued for URLs without the hash.¹⁷ That’s dated advice from a practitioner, not a Baidu rule.

Use server redirects and not script redirects, given how Baidu lists JS redirects.³¹⁸

How to test what Baidu gets from your site

Start with the checks that need no account, then move to Baidu’s own tool. The first two steps take minutes.

1. View the source. Open view-source: on a key page, not the inspector. Search for your H1, a sentence of body text and your main links.

2. Fetch it without a browser. A plain request runs no JavaScript, so it shows the first response. This one uses Baidu’s published user agent:¹

   curl -s -A "Mozilla/5.0 (compatible; Baiduspider-render/2.0; +http://www.baidu.com/search/spider.html)" https://example.com/page/ | wc -c

   Drop | wc -c to read the HTML. Keep the pipe to count its size in bytes, for the 128k check.

3. Run crawl diagnosis. The tool shows what the spider fetched, for PC and mobile.⁷³² It allows 70 uses per site per week and shows the first 200KB.⁷ Baidu cut the quota from 200 in July 2020, and said the tool tests crawling and doesn’t affect indexing.²⁹

4. Check your logs. Look for Baiduspider-render, then confirm the IP by reverse DNS. Real hosts end in .baidu.com or .baidu.jp.¹⁹ A hit shows a fetch, not that the content was indexed.

The tool sits under Baidu’s site resources menu,²⁴ and you need a verified site. Our guide to setting up Baidu Webmaster Tools covers that.

Run your own marker test

You can’t view Baidu’s stored copy anymore, since its snapshot feature disappeared in August 2022.²³ And site: is only an estimate.²⁵

A marker test is the closest thing to a direct answer. We haven’t run this one ourselves yet.

1. Build one test page. Use your normal template, with enough real content to be worth indexing.
2. Add two made-up strings. Put one in the HTML, and have a script insert the other after load. Neither should exist elsewhere online.
3. Link and submit the page. Baidu says submission doesn’t guarantee indexing,²⁶ and it gives no timetable.²⁷
4. Wait for the page to appear. Our guide on checking which pages Baidu has indexed shows how.
5. Search for each string. If the HTML string appears and the script string doesn’t, Baidu hasn’t indexed the injected text on that page.

The result covers one page, on one date. If rendering is selective, as Botify claimed in 2019,¹¹ your next page may behave differently.

If a page is missing entirely, JavaScript may not be the cause. Start with why your site isn’t showing on Baidu.

For phones, see Baidu mobile SEO as well.

The bottom line

Baidu has had a rendering agent since 2017, and we found no statement of how much it does. The two controlled tests that reached a result show Baidu missing injected content, though neither is recent.

So treat JavaScript as a bonus, not a delivery method. If your text and links are in the first HTML response, the question stops mattering.

How Index in China can help

We translate your site into Simplified Chinese with a human reviewer, handle its submission to Chinese search engines where the account allows it and send a monthly report on Chinese traffic.

Tell us early if your site depends on client-side JavaScript, because that shapes the work.

We don’t promise rankings or traffic. See what we do.

Book a 30-minute call

Monthly report BaiduIndexed Bing ChinaIndexed SogouIndexed