FAQ
Last updated: 26 October 2026.
Most requests to InfraRabbit are handled by automated flows. This page says what each flow does, so you know what will happen before you write. We do not promise a reply time for any mailbox. The flows act when the request arrives, not when someone reads it.
Support
How do I get help?
Read this page first. If it does not answer your question, write to support@infrarabbit.com. You get an automatic reply with a link to this page. Mail that no rule recognises is read later, in a periodic review, by the operator. We do not promise an answer or a time.
Is there a phone number or chat?
No.
Which mailbox is for what?
| Address | Use it for | What happens |
|---|---|---|
| corrections@infrarabbit.com | A wrong price, a wrong crawler verdict, a privacy request | Acknowledged automatically. A price correction triggers a re-fetch of that source (see below) |
| crawlers@infrarabbit.com | A complaint about InfraRabbitBot, or a request to exclude your site | The sender is checked. A verified request adds an exclusion (see below) |
| abuse@infrarabbit.com | Abuse reports about our servers or crawler | Same flow as crawlers@ |
| press@infrarabbit.com | Press questions | Automatic reply with the press kit link |
| security@infrarabbit.com | A security problem | See Security |
| support@infrarabbit.com | Everything else | Automatic reply with this page |
Corrections
A price is wrong. What do I do?
Write to corrections@infrarabbit.com. Name the SKU or series, the value you expect, and a link to the provider's own page. If you can, send the URL of our page.
What happens, automatically:
- The mail is acknowledged.
- The source for that SKU is queued for a re-fetch. The ledger reads the provider's page again.
- A diff note is attached: what the page said before, what it says now.
- The correction is logged publicly, with its date. If the provider's page and our price already agree, the log says so and the price stays.
The original record stays visible. A correction is a new dated entry, not an edit (charter rule 2 and 10).
If our number matches the provider's page but you think the provider is wrong, that is not a correction for us. We publish what the provider publishes.
A crawler verdict is wrong.
Write to corrections@infrarabbit.com with the domain and agent. Check first: the verdict is based on the robots.txt we fetched on the date shown. If your robots.txt changed after that, the new file shows in the next daily run. Observations already sealed in a daily proof do not change. A verdict can still be wrong; every verdict links to the stored robots.txt body and to the rule we applied. The parser is documented on /crawlers/methodology.
Exclusions
How do I have my website excluded from the crawler archive?
Prove that you control the domain in one of three ways. The request is then applied automatically.
- DNS. Add a TXT record on the domain that starts with
infrarabbit-exclude=and has any non-empty value, for exampleinfrarabbit-exclude=yes. - robots.txt. Add
User-agent: InfraRabbitBotandDisallow: /to your robots.txt, and write to crawlers@infrarabbit.com. (The Disallow already stops every fetch except robots.txt itself.) - Email from the domain. Write to crawlers@infrarabbit.com from an address at the domain.
What happens when a request is verified:
- The domain is added to the exclusion list. An exclusion of
example.comcovers its subdomains. A request for one subdomain covers only that host. - The site is skipped in the daily run, in backfill and in live lookups.
- It is shown as "excluded at the owner's request on <date>", and left out of shares and denominators. The number of excluded sites is published.
- You get a fixed reply that says the exclusion is in place.
Observations sealed before your request stay in the daily proofs, because a sealed proof cannot be changed without breaking it. They are no longer shown or counted.
A request we cannot verify is not applied automatically. To finish it, use method 1 or 2. The fetcher page at /crawlers/fetcher has the details and what the crawler fetches.
I got a complaint-style mail from my hosting provider about InfraRabbitBot.
Same flow. A complaint forwarded by Hetzner's abuse desk excludes the domain on the first complaint. Other complaints exclude the domain when the sender is at that domain or the domain proves control by DNS. A second complaint within 30 days automatically halves the crawler's request rate and pauses the random-path probe. Our fetcher makes about six requests per site per day (the daily robots.txt fetch with any redirects, up to three more files and the homepage where robots.txt allows it, and one check that tells real files from catch-all pages) and at most one request per second per host. It identifies itself as InfraRabbitBot.
Takedowns
A takedown request for other material.
If you think we show material that infringes your rights (for example a quoted text or a logo), write to corrections@infrarabbit.com. Say what the material is, where it is, and why you hold the right. The request is read by the operator. We do not promise a reply time.
Privacy requests
How do I ask what you hold about me, or have it deleted?
See Privacy, "Your rights". Write to corrections@infrarabbit.com. We usually hold very little: access logs for 7 days and some hashed values.
Data and API
Can I use your data?
Yes. Our measurements are CC BY 4.0: name InfraRabbit, link, state the data date. Raw AI text and text from other sources is not licensed. See the licence matrix.
How do I cite you?
Every page has a "Cite this" box. The format is on /methodology.
Is the API free?
The free API needs no key. Pro and Team plans add per-SKU history beyond 90 days, bulk exports, alerts and higher limits. See /pricing.
What happens if I go over the rate limit?
You get HTTP 429 with a Retry-After header. Wait that long. See /api.
Will the API change?
/v1 is frozen. We give 6 months' notice before any breaking change.
About InfraRabbit
Are you affiliated with any provider?
No. InfraRabbit is independent. It is not owned by, paid by or affiliated with any provider, AI company or website it measures. We do not take affiliate money for rankings. See the Charter.
Why is a provider missing?
Some providers block automated access or load prices with JavaScript that we cannot read deterministically. As of 26 Oct 2026, Contabo is not covered: its site blocks automated access, and we do not get around blocks. The methodology page lists what we do not have and the date it was written. "Cheapest" means the cheapest among the providers we read, not in the market.
Do you use AI to measure?
No. Prices and crawler verdicts come from deterministic code. No language model produces a number we publish.