Googlebot renders JavaScript. OpenAI, Anthropic and Perplexity don't say their search crawlers do. See what each doc says and how to check your site.
The short answer
Googlebot runs JavaScript. OpenAI, Anthropic and Perplexity each publish a page about their search crawler, and none of those pages says the crawler runs your JavaScript.
That gap decides what you should do. If your words only appear after a script runs, Google can still find them once it renders the page. The crawlers behind ChatGPT search, Claude and Perplexity may find an empty page. If your words are already in the HTML your server sends, every reader gets them, whether it runs scripts or not.
The rest of this post shows where each claim comes from and how to check your own site in a few minutes.
What Google says about JavaScript
Google is the one crawler with a detailed public answer. Its [JavaScript SEO basics](https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics) guide says Google Search runs JavaScript with an evergreen version of Chromium.
It works in three phases: crawling, rendering and indexing. Googlebot fetches your URL and reads the HTML. It then queues the page for rendering, where a headless Chromium runs your scripts. Google indexes the rendered HTML. The guide says a page may wait in the render queue for a few seconds, and sometimes longer.
The guide also describes the app shell model, where the first HTML holds no content and Google has to execute JavaScript to see the page. Google handles that. Even so, it recommends server-side or pre-rendering, partly for speed and partly because, in the guide's words, "not all bots can run JavaScript."
What the AI search crawlers document
ChatGPT search, Claude and Perplexity each run their own search crawler. Here is what each vendor says about it.
**OAI-SearchBot (ChatGPT search).** [OpenAI's crawler page](https://platform.openai.com/docs/bots) says OAI-SearchBot is used to surface websites in search results in ChatGPT's search features. Sites that opt out won't be shown in ChatGPT search answers, though they can still appear as navigational links.
**Claude-SearchBot (Claude).** [Anthropic's help article on its crawlers](https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler) says Claude-SearchBot exists to improve the quality of search results. Disabling it may reduce your visibility in Claude's search results.
**PerplexityBot (Perplexity).** [Perplexity's crawler docs](https://docs.perplexity.ai/guides/bots) say PerplexityBot is designed to surface and link websites in Perplexity search results.
Each vendor also lists a second bot that visits a page when a person asks about it: ChatGPT-User, Claude-User and Perplexity-User. Those pages are just as quiet on JavaScript.
Look at what the three pages cover: what each bot is for and how to allow or block it in robots.txt. They never mention JavaScript. They don't say the bots run it, and they don't say the bots can't.
Treat that as a reason to be careful, not as proof that these crawlers can't render. Some may well run scripts today. But behavior nobody has documented can change without notice, and you can't plan around it. What is documented is that these bots fetch your page. The HTML in that response is what you can count on.
Why server-rendered HTML is the reliable choice
A client-rendered React app, the kind a default Vite setup produces, sends HTML like this:
```html <!DOCTYPE html> <html lang="en"> <head> <title>My Awesome App</title> </head> <body> <div id="root"></div> <script type="module" src="/src/main.tsx"></script> </body> </html> ```
That empty `<div id="root"></div>` is the whole page until a script builds the rest in the browser. Googlebot will run the script. A reader that doesn't run it sees an empty container. A reader that does run it can still lose your words if a script fails before it paints them.
Server rendering (SSR) builds the HTML for each request. Static generation builds it once, at build time. Either way, your words are in the first response, before any JavaScript runs. Frameworks like Next.js, Astro, Nuxt and TanStack Start do this out of the box.
You don't have to rewrite a React app to get there. A client-rendered app can be pre-rendered at build time, so each route ships as a real HTML file. If you host with cite-met, we run your own build and serve what it writes, so set up pre-rendering in your build and you can keep your React code.
One more reader is easy to forget. Link previews in chat and social apps are built from the [Open Graph tags](https://ogp.me) in your HTML: og:title, og:description and og:image. If a script adds those tags, a preview tool that doesn't run scripts won't see them, and your shared link looks bare. That has nothing to do with AI search, but the fix is the same.
How to check your own site
You don't have to guess what your server sends. Two checks take a few minutes.
**Run the scan.** The [free scan at cite-met.com](https://cite-met.com) fetches your homepage once and counts the readable words in the HTML. Fewer than 80 words inside an empty app container, like `<div id="root"></div>`, fails the check. It then runs the same test on up to three inner pages, so a site that only renders its homepage properly doesn't get a pass. The scan also checks whether your robots.txt lets OAI-SearchBot, Claude-SearchBot and PerplexityBot in.
**Ask your server directly.** Pick a sentence from your page and look for it in what the server sends back:
```bash curl -sL https://example.com/ | grep -i "a short phrase from your page" ```
If grep prints a line, your words are in the HTML. If it prints nothing, they arrive after JavaScript runs. Use a short phrase with no apostrophes or special characters, because HTML often encodes those. Then repeat the check on two or three inner pages.
In a browser, View Source shows the HTML as the server sent it. The Elements panel in developer tools shows the page after JavaScript ran, so it can look fine when the HTML is empty. To see what Google rendered, the JavaScript SEO guide points to the Rich Results Test and the URL Inspection Tool.
Questions
**Does ChatGPT run JavaScript?** OpenAI's crawler page doesn't say. It describes OAI-SearchBot (ChatGPT search), GPTBot (model training) and ChatGPT-User (visits a person triggers), and it says nothing about running scripts for any of them. Put your words in the HTML and the question stops mattering.
**Do Claude and Perplexity run JavaScript?** Anthropic's help article and Perplexity's crawler docs don't say either. Both describe their bots (Claude-SearchBot and PerplexityBot for search, Claude-User and Perplexity-User for visits a person triggers) and how to control them in robots.txt. Neither describes script execution.
**If Google renders JavaScript, why bother with server rendering?** Rendering is a second step. Google's guide says pages wait in a render queue, sometimes longer than a few seconds, and that not every bot runs JavaScript. Words that are already in the HTML are indexed without that wait, and a script error can't remove them. They also reach the readers that never run scripts.
Sources
We read these pages on 15 October 2026. Vendors change their crawler pages, so check the current version before you rely on any detail here.
* [Google: Understand JavaScript SEO basics](https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics) * [OpenAI: Overview of OpenAI crawlers](https://platform.openai.com/docs/bots) * [Anthropic: Does Anthropic crawl data from the web, and how can site owners block the crawler?](https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler) * [Perplexity: Perplexity crawlers](https://docs.perplexity.ai/guides/bots) * [The Open Graph protocol](https://ogp.me)
Scanner checks this post covers
- CM-101 Content in the HTML: Content in the HTML means your homepage's words arrive in the first response from your server, before any JavaScript runs.
- CM-104 Search and AI search crawlers allowed: Search crawlers allowed means your robots.txt lets Googlebot, Bingbot, OAI-SearchBot, Claude-SearchBot and PerplexityBot fetch your pages.