No. 27 of the most-visited websites in Germany · As of November 15, 2026
Can ChatGPT, Claude and Gemini read WELT?
Every morning, we open the welt.de homepage with the same fetch tools ChatGPT, Claude and Gemini use. This page shows which AI can read it, which gets turned away and why.
The short answer
ChatGPT can read welt.de: its fetch tool opened the homepage. Claude cannot read welt.de, and the website isn’t the reason: Anthropic itself refuses the fetch (“url_not_allowed”). welt.de’s robots.txt allows Claude-User. Gemini uses welt.de as a source, but through Google Search. Google’s own fetch tool rejected the homepage. Meta AI: robots.txt allows Meta’s fetcher. We can’t test it, because Meta offers no public fetch tool.
Result per AI
What welt.de’s robots.txt says
User-agent: * Disallow: /channels-extern/ Disallow: /sportdaten/ Disallow: /testgpr/ Disallow: /boerse/data/ Disallow: /partner/ Disallow: /reisetipps/ Disallow: /z/ Disallow: /appl/ Disallow: /woa/ Disallow: /am-sonntag/vorproduktion/ Disallow: /audiofiles/ Disallow: /out-of-home/ Disallow: /immobilien/expose Disallow: /suche Disallow: /onward/ Disallow: /api/ Disallow: /*?config Disallow: /*?config=newsmli_bloomberg2 Disallow: /*.xmli Disallow: /*?service=ajax Disallow: /*?service=Ajax Disallow: /*?ajax Disallow: /*?ajax&wid Disallow: /*?config=print Disallow: /*?config=articleidfromurl Disallow: /*?config=endscreen Disallow: /*?config=iframewelt Disallow: /*?config=langeslestueck Disallow: /*?config=latest_videos Disallow: /*?config=menu_home Disallow: /*?config=mostviewed_videos Disallow: /*?config=recommended_videos Disallow: /*?config=regioarticlemarginal Disallow: /*?config=zoom Disallow: /*?config=zoomopener Disallow: /*?noredirect=true&config=standalone Disallow: /*?config=standalone Disallow: /*?wtmc=XING Disallow: /*?config=articleidfromurl Disallow: /*?print=yes Disallow: /*?tabPane Disallow: /video/embeded/ Disallow: /img/*-wWIDTH*.jpg Disallow: /assets/public/fonts/
Excerpt showing only the rules for AI agents and for all bots (*). robots.txt is a request from the website, not a technical lock. Anthropic says its bots honor it; for the ChatGPT-User fetcher, OpenAI says the rules may not apply.
Bot protection
welt.de uses bot protection. An AI agent working for you in a browser may run into a captcha there.
Changes
Nothing has changed for welt.de since daily checks began on November 14, 2026.
Frequently asked questions about WELT
Can ChatGPT read welt.de?
Yes. ChatGPT can read welt.de: its fetch tool opened the homepage. (As of November 15, 2026)
Can Claude read welt.de?
No. Claude cannot read welt.de, and the website isn’t the reason: Anthropic itself refuses the fetch (“url_not_allowed”). welt.de’s robots.txt allows Claude-User. (As of November 15, 2026)
Can Gemini read welt.de?
Yes. Gemini uses welt.de as a source, but through Google Search. Google’s own fetch tool rejected the homepage. (As of November 15, 2026)
Does WELT block AI crawlers in robots.txt?
No. Its robots.txt blocks none of the AI names used by OpenAI, Anthropic, Google and Meta.
How to cite this data
Source: i4u AI Agent Radar, i4u.business/en/ki-agent-radar/welt-de/
Data as of: November 15, 2026 · Licence: CC BY 4.0, credit “i4u” with a link. · Press and questions: info@i4u.business
Download the data: CSV, Top 50 · JSON, Top 50