Skip to content
All articles
AI search

Test: ChatGPT and Claude still miss JavaScript content

A new test found that three AI assistants saw only the menu and footer of pages built with JavaScript. Here is what the crawler docs say, and how to check your site.

S
SEOSFT · Product team
4 min read

If JavaScript adds the main text of your pages in the browser, some AI assistants may see only your menu and footer. That is the result of a test that SEO consultant Glenn Gabe published on October 19, 2026. It is a useful check for any business that wants to appear in AI answers.

What the test found

In August 2025, Gabe tested whether AI search tools could read pages whose content JavaScript renders. Most could not. 434 days later he ran the test again. He asked ChatGPT, Claude, Perplexity, Gemini and Copilot to read each page and give its first and last paragraphs.

He tested three kinds of pages:

  • A page with all of its content rendered by JavaScript. ChatGPT, Claude and Perplexity failed. ChatGPT said its page extraction "returned only navigation and footer elements, not the main body text." Gemini and Copilot read the content.
  • A page with a normal HTML header and footer, but the main content in JavaScript. Same result: ChatGPT, Claude and Perplexity missed the content. Gemini and Copilot read it.
  • A parallax site built with WebGL graphics. ChatGPT, Claude, Perplexity and Copilot returned only the page's "content did not load" error message. Only Gemini read the content.

Gabe explains that Gemini uses Google's search systems and Copilot uses Bing's. His conclusion: "over a year later, AI Search platforms still cannot render JavaScript-based content."

This is one expert's test on three pages. It is not a statement from OpenAI, Anthropic or Perplexity. But the results match his 2025 test, and the assistants themselves gave the reason. Claude, for example, said its plain HTML fetch does not run JavaScript.

Which crawlers run JavaScript

Gabe asked the assistants to fetch pages in a chat. In a chat, the vendors use their user-triggered agents: ChatGPT-User, Claude-User and Perplexity-User. Their search crawlers (OAI-SearchBot, Claude-SearchBot and PerplexityBot) are separate bots. The crawler documentation of OpenAI, Anthropic and Perplexity lists these bots and their purposes. It does not say whether any of them runs JavaScript.

Only two of the large crawlers document JavaScript support:

  • Googlebot renders pages with an evergreen version of Chromium, says Google Search Central. Google still adds: "server-side or pre-rendering is still a great idea because it makes your website faster for users and crawlers, and not all bots can run JavaScript."
  • Bingbot is "generally able to render JavaScript," Bing said in 2018. Bing also said it cannot do this at scale on every page, and it recommended prerendered HTML for sites that depend on JavaScript.

Why it matters

Some sites, often single-page apps, add their text in the browser after the page arrives. Google can usually process that. But if an AI assistant gets only an empty shell, it cannot quote, summarize or recommend what is on the page.

For a shop, product details can be invisible. For a service business, it can be the page that explains what you do.

What to do

  • Look at the page source. Open a key page and view its source (Ctrl+U on Windows, Cmd+Option+U on a Mac in Chrome). Search for a sentence from the main text. If it is not there, JavaScript adds it.
  • Test it the way Gabe did. Ask ChatGPT or Claude for the first and last paragraphs of the page. Do not only ask whether it can read the page. Compare the text it returns with the page.
  • Serve important content in the HTML. Ask your developer about server-side rendering or static generation for key pages: homepage, services, products and pricing.
  • Give animated pages a text version. If you use a heavily animated design, put the core message in the HTML as plain text too.
  • Let the AI search crawlers in. Rendered HTML does not help if robots.txt blocks OAI-SearchBot, Claude-SearchBot or PerplexityBot.
The SEOSFT AI search check reads your robots.txt and shows, bot by bot, which AI search crawlers can reach your site. The sample report at seosft.com/demo shows what an audit checks.

Sources

  • Glenn Gabe (GSQI), 434 Days Later – Are AI Search Platforms Like ChatGPT and Claude Still Unable To Render JavaScript In October 2026? (October 19, 2026): https://www.gsqi.com/marketing-blog/can-ai-search-platforms-render-javascript/
  • Glenn Gabe (GSQI), AI Search and JavaScript Rendering (2025 case study): https://www.gsqi.com/marketing-blog/ai-search-javascript-rendering/
  • Google Search Central, Understand the JavaScript SEO basics (updated March 4, 2026): https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics
  • Bing Webmaster Blog, JavaScript, Dynamic Rendering, and Cloaking (October 2018): https://blogs.bing.com/webmaster/october-2018/bingbot-Series-JavaScript,-Dynamic-Rendering,-and-Cloaking-Oh-My
  • OpenAI, Overview of OpenAI crawlers: https://developers.openai.com/api/docs/bots
  • Anthropic, Does Anthropic crawl data from the web?: https://support.claude.com/en/articles/8896518
  • Perplexity, Perplexity crawlers: https://docs.perplexity.ai/guides/bots

Check this on your own site

Free audit: every page of your site and a ranked plan.

More guides

Core Web Vitals fixes: LCP, INP and CLS in plain English

9 min read · October 21, 2026

Google adds more links to AI Mode and AI Overviews

3 min read · October 19, 2026

Which AI crawlers to allow in robots.txt

5 min read · October 2, 2026

llms.txt: what it does, and what it does not

4 min read · September 28, 2026