General settings

Settings > General is a short tab: schema, llms.txt, and two API keys.

Auto-generate schema

On by default. When you save a post, SEOBetter generates JSON-LD schema (structured data) for it.

Straight talk on who actually reads schema: Google uses it for rich results (star ratings, FAQ dropdowns, recipe cards) but has said its AI Overviews don't rely on it — there, the content itself does the work. Bing (which also powers Copilot and ChatGPT browsing) is a documented schema consumer, and Perplexity parses structured data when citing sources. So schema won't buy you a Google AI Overview by itself, but it wins rich results and helps every non-Google engine understand your pages. It costs nothing at page load, so leave it on.

If you run Yoast, Rank Math, or AIOSEO, SEOBetter defers to their schema output where they already cover it, so you won't get duplicates.

llms.txt

On by default. This publishes a file at yoursite.com/llms.txt, a machine-readable index AI systems can use to find your best content. On the free plan it's a flat list of your 20 most recent posts.

Honest status: llms.txt is an emerging convention, not an accepted standard. Google has said it doesn't use it, and no major AI company has officially committed to reading it — though AI crawlers are regularly observed fetching the file from sites that publish one. Treat it as zero-cost insurance for whatever reads it next, not a ranking lever. It's one static file, so there's no downside to leaving it on.

Pro adds content-type categorization and a quality filter (only posts with a GEO score of 40 or more are listed), plus a custom one-line site summary you can edit here. Pro+ adds /llms-full.txt and per-language variants.

The file regenerates automatically every time you save a post. Pro users also get a "Regenerate now" button for manual refreshes.

Which engine reads what

A quick honesty map, so you know what each setting actually buys you:

• Google Search + AI Overviews — reads your content and classic SEO signals. Schema earns rich results (not AI Overviews); llms.txt is ignored.
• Bing + Copilot — documented schema consumer. Bing hygiene matters double because ChatGPT's browsing runs on Bing's index.
• Perplexity — crawls aggressively and cites sources heavily; parses structured data. Make sure PerplexityBot is allowed (the free AI Crawler Audit checks this).
• Claude / ChatGPT crawlers — visibility starts with robots.txt access (GPTBot, ClaudeBot). The AI Crawler Audit tells you exactly who's blocked.

The common thread: well-structured, well-sourced content is the one signal every engine uses. That's what the generator and GEO score optimize for.

One important caveat on crawler access: the AI Crawler Audit reads your WordPress origin — robots.txt, meta robots and the X-Robots-Tag header. A CDN or firewall sitting in front of your site can block the same bots upstream, where WordPress never sees the request, and the audit will still show "Allowed". Cloudflare is the common case: its AI Crawl Control page has per-bot Block toggles (and Bot Fight Mode can sweep AI crawlers up too) — we've seen a production site block every AI crawler and Googlebot there while robots.txt said everything was allowed. If a bot shows Allowed in the audit but never appears in your server logs or AI search results, check your CDN dashboard first.

Tavily API key (Pro)

Tavily is a search API that powers real expert quotes and citations with verified source URLs. It extracts actual text from real web pages, so quotes are verbatim rather than made up.

This field is Pro-locked. If you're on Pro, get a free key at tavily.com (their free tier covers 1,000 searches a month) and paste it here. You supply the key; SEOBetter unlocks the feature.

Without Tavily, articles still get citations from the default research sources. Tavily adds the verbatim-quote layer on top.

Pexels API key

Available on every plan and strongly recommended. Your own free Pexels key (pexels.com/api, 15,000 requests a month, no card) gives every article on-topic stock photos with a dedicated quota. Without one, images come from a shared cloud pool with a small monthly allowance, then generic Openverse images. See Add a free Pexels key for the walkthrough.

About the masked key fields

Saved keys display as dots. Leaving a key field blank and saving keeps the existing key. To actually delete a key, tick its "Remove saved key" checkbox before saving.