Audit your site for SEO and AI search.
Crawlwise crawls up to 25 pages, checks what search engines and AI assistants need to find and understand them, and tells you which fixes matter most.
- A score out of 100
- Titles, descriptions, headings, image alt text and thin pages, measured on every page.
- AI search access
- Whether ChatGPT, Claude, Perplexity and Gemini are allowed to read your site.
- Five fixes, in order
- Ranked by impact, with a prompt you can hand straight to your coding agent.
- Thin content on 9 pages
- 6 titles outside 30–60 characters
- Heading hierarchy skips a level
How it works
Three steps, no setup, no plugin.
Drop in a URL
We crawl up to 25 pages and pull titles, meta descriptions, headings, canonicals, structured data, word counts and alt text — at one request per second, so your server barely notices.
The model reads the pages
Not just the tags — the copy. It infers what you sell, who you sell to, and which search intent each page is actually trying to serve.
You get a ranked plan
Five things to do next, each with the reason, the effort, the expected impact, and example copy you can paste in.
Generic advice is why nothing moved last quarter.
Every other tool tells you "add a meta description." Crawlwise writes the one for /extensions/submit based on what that page is for.
⚠ Meta description length not optimal — 5 pages affected.
This page is a submission form for community extensions, so the description should reduce friction, not describe the product. Try: "Submit your plugin and make it available to the community — review usually takes 48 hours." (90 chars)
Per-page intent analysis
For each key page: the intent it serves, what is missing to fully satisfy it, and four concrete additions.
A plan, in order
Five things to do next, ranked by impact against effort — with the reason, the affected pages and copy you can paste in.
Nothing invented
A page read that names a URL we never fetched is thrown away before it reaches your report. Every claim is anchored to crawled evidence.
Questions
What is Crawlwise?
Crawlwise is an AI SEO audit tool. It crawls up to 25 pages of a public website, reads each one with a language model, and returns a ranked list of what to change — with the example copy to change it to. An audit takes about two minutes and needs no account.
Is Crawlwise free? Do I need an account?
Crawlwise is free and needs no account and no card. Technical audits are unlimited, and 3 full AI audits run per day.
Do I need to install anything on my site?
No. We crawl publicly reachable pages the same way a search engine does — no plugin, no script, no account. Pages behind a login are not reachable and are skipped.
How is this different from a classic SEO crawler?
A crawler compares your tags to generic rules. Crawlwise reads the actual copy, works out what each page is for, and writes advice specific to your product — including the sentences to use.
Will this slow my site down?
No. We read robots.txt before the first page, honour Crawl-delay, and space requests one second apart per host — concurrent audits of the same site share that pace rather than doubling it. A 25-page crawl is about 30 seconds of light traffic.
What does the audit not cover yet?
Competitor comparison, scheduled re-scans, keyword positions and regression alerts are not built. Neither is JavaScript rendering — a page that ships its copy only from the client will read as thin.
Does the AI make things up?
Every recommendation is anchored to a specific URL and the crawled evidence behind it — the page count, the measured value, the exact tag. If we can’t measure it, we don’t claim it.
How long does an audit take?
A Crawlwise audit takes about two minutes for a 25-page site. The crawl itself is roughly 30 seconds, because requests are spaced 1 second apart per host; the rest is the language model reading the pages.
What does Crawlwise check?
Crawlwise measures titles, headings, meta descriptions, structured data types and content depth on every page it reads. Each is judged against what the page is actually for rather than against a generic rule — a thin title on a pricing page and a thin title on an article are not the same problem.
Can AI assistants like ChatGPT and Claude cite my site?
Crawlwise reports whether an answer engine could quote each page: whether the page opens with a self-contained definition of its subject, whether its answers survive being quoted out of context, and whether it offers lists or tables an engine can lift. It also checks robots.txt for GPTBot, ClaudeBot, PerplexityBot, Google-Extended and OAI-SearchBot, and flags a missing /llms.txt.
Can I hand the report to a coding agent?
Every Crawlwise report includes a prompt written for a coding agent. It carries the findings, the URLs and the evidence behind each one, so an agent can act on the report directly instead of you retyping it.
What data does Crawlwise store?
Crawlwise stores the measurements taken from your pages, not the text of them. It does not collect email addresses, and does not resell what it crawls.
How do I stop Crawlwise crawling my site?
Two lines in robots.txt stop CrawlwiseBot: a User-agent line naming CrawlwiseBot, and Disallow followed by a slash. CrawlwiseBot reads robots.txt before requesting any page and obeys it — no account and no request to us needed. About CrawlwiseBot