A free Website AI Score account includes 10 credits, and 10 credits is enough to run a complete first cycle of AI visibility work without paying anything. The free scan itself needs no account at all: it rates any URL across six LLMs and lists what is broken. The credits are what turn the diagnosis into the fix. One credit diagnoses a page's specific content gaps against a target phrase; four credits generates a publish-ready article, rewrites an existing one with the facts preserved, or produces a CMS-formatted landing section. So 10 free credits is, for example, two diagnoses plus two generations, or ten diagnoses, or a diagnose-rewrite-confirm loop on your most valuable page with credits left over. Here is exactly what to spend them on.
The word "free" in AI tools usually means a demo, a watered-down tier, or a trial that expires. This is worth stating plainly because it is different: the scan is free without an account, permanently, and the 10 credits on registration are the same credits paid users spend, on the same tools, producing the same output. The credits are designed to move a page through a diagnostic workflow, scan, diagnose, generate or rewrite, re-scan, not to produce more AI text. This page explains that workflow; the linked articles explain the methodology behind each step.
Scan, diagnose, fix, confirm: that first cycle is what the credits are for. Ten covers it on one important page with room to spare, and running it once teaches you more about your AI visibility than any dashboard.
What is free without an account?
The scan. Paste any URL on the homepage and it runs the full 10-check audit, scores the page across GPT, Claude, Gemini, Llama, Perplexity, and Grok, and returns what is broken with a fix category for each issue. No card, no signup, no limit that matters for a first look. It is the same audit the paid version uses, which means the free result is a real diagnosis, not a teaser. If the scan shows your page renders as an empty shell, you know your first fix is a developer's, and you knew it for nothing. How to read the scan result covers what each finding means. Run it on your three most important pages before you register, because the scan tells you which pages deserve the credits.
What do the 10 credits buy?
Access to the Content Creator, priced per action. Diagnose Page costs one credit and returns a page's ranked content gaps against a target phrase. The four generation modes cost four credits each: Direct Generation writes an article from a topic, Blog from Gap writes the article that closes a diagnosed gap, Rewrite restructures an existing article with every fact verified, and On-Page Section produces a landing or product section formatted for your CMS. Every generation ships with schema, meta, and slug. And every mode refunds the credit if the output fails a quality check, so a failed run costs nothing. Ten credits is ten diagnoses, or two generations and two diagnoses, or one generation and six diagnoses, depending on what the scan told you.
How should you spend them?
On one page, end to end, before spreading them thin. Pick the page the free scan showed as readable but poorly cited, usually your most valuable commercial or cornerstone page. Diagnose it against the phrase it should win: one credit, and you now know the exact sub-questions it fails on. Fix the top gap: four credits for a Rewrite that restructures the page and folds in the missing angle with every fact preserved, or for a Blog from Gap article if the gap deserves its own page. Re-diagnose: one credit to confirm the gap closed. That is six credits for a measured improvement on the page that matters most, with four left to diagnose the next candidates. The alternative, generating two articles on topics you guessed at, spends eight credits and teaches you nothing.
What should you not spend them on?
Generating before diagnosing, which is the default mistake and the reason most AI content never gets cited: the writer guesses at coverage and guesses wrong in the same way every time. Diagnosing a page the free scan already showed is unreadable, because content gaps on an empty shell are irrelevant until the rendering is fixed, which is a developer's job the scan already flagged. And diagnosing pages that do not matter, since a diagnosis on a page you will never fix is a credit spent on curiosity. The scan is free precisely so you can be curious there and spend the credits where they change something.
What happens after the free credits?
The same tools at $49 for 100 credits, which is 100 diagnoses, or 25 generated articles, or any mix, roughly two dollars per publish-ready piece with schema attached. Credits are bought as packs, not a subscription, and the account and the free scan stay available either way. The first cycle on free credits is meant to show you whether the loop moves your citations before you decide whether to buy the next hundred, and if the re-scan after your first rewrite shows nothing moved, you have lost nothing and learned something. Register, spend the ten on one page done properly, and let the result make the decision.
Spent as a sampler, a few clicks to see what the tool does, free credits show you what it does and nothing about whether it works for you. Spent as a cycle on one real page, they answer the only question worth asking before paying for anything: did the citations move. Ten credits is enough to find out, which is more than most trials offer and more than most people bother to test.
Sources
- Website AI Score, free scan: the 10-check, six-LLM audit with no account. websiteaiscore.com
- Website AI Score, Content Creator: the five modes and per-action credit costs. websiteaiscore.com/content-creator
- Website AI Score, content gap diagnosis: what the one-credit diagnosis returns. View article
- Website AI Score, rewrite without losing facts: what the four-credit rewrite does and verifies. View article
- Website AI Score, rate my website for AI: how to read the free scan result. View article
