A bot reads your article. You pay for the reporting, the editor and the hosting. Someone else gets the information. Charging for that visit sounds reasonable. The awkward question is whether the visitor will buy access or simply leave.
Cloudflare Pay Per Crawl makes that transaction possible. It does not make the buyer appear. For a small publisher, that distinction matters more than the button in the dashboard.
As checked on 28 September 2026, Cloudflare still describes Pay Per Crawl as a closed beta. Publishers can apply; existing Enterprise customers can contact their account executive. This is a review of the documented mechanics, not an AntiMedia earnings case study. Cloudflare's feature overview explains the current access model.
What does Cloudflare Pay Per Crawl actually sell?
It sells crawler access to content. A participating crawler can indicate that it intends to pay and retrieve a page successfully. Otherwise, it can receive an HTTP 402 Payment Required response with pricing information. That payment request is not itself a sale.
Cloudflare handles the transaction infrastructure as Merchant of Record. It also warns that WAF or Bot Management blocks override the charge setting: a crawler you have blocked there cannot reach the content just because you have set a price. These mechanics are documented in the Pay Per Crawl overview.
So the first useful question is not "How many AI bots visit my site?" It is "How many requests turn into paid access?" A large crawler count can look promising while producing no money at all.
Can a small website enable it?
Being small is not the same thing as being eligible. Start by confirming beta access, then check the account requirements.
Cloudflare's setup instructions call for an active account, a domain using Cloudflare nameservers or Cloudflare-managed DNS, and Administrator or Super Administrator permissions for account-level configuration. Making the controls visible for a domain does not itself change its security rules or start earning money.
Before moving infrastructure for this feature, ask Cloudflare whether your site can join and whether your business can complete payout onboarding. Do not turn a speculative revenue test into a migration project first.
How much could a publisher earn?
Cloudflare currently lists a minimum price of $0.001 per crawl. Its pricing instructions define the charge around successful content retrieval with an HTTP 200 response and also describe optional dynamic pricing.
Here is the basic arithmetic at that minimum price. These are hypothetical volumes, not observed publisher results or a forecast.
| Paid requests | Gross charges at $0.001 each |
|---|---|
| 1,000 | $1 |
| 10,000 | $10 |
| 100,000 | $100 |
| 1,000,000 | $1,000 |
The word "paid" does all the work in that table. A million bot requests in your logs is not a million purchases. Nor is the gross total a statement of what lands in your bank account.
Raising the price changes the arithmetic, but you cannot assume the same buyers will keep requesting the same pages. A sensible test measures demand at the chosen price instead of multiplying yesterday's free crawls by tomorrow's fee.
Getting paid has its own conditions
Cloudflare requires a dedicated Stripe Connect account created through its dashboard; an existing Stripe account is not compatible with this feature. Its payout documentation describes monthly payments to publishers in good standing, subject to settlement periods and minimum payout thresholds.
There is another practical limitation: the documentation says the accrued balance is not currently visible in the dashboard. Publishers can request updates from their Cloudflare team. That page does not specify the threshold amount or a complete fee schedule, so get those terms confirmed before treating a gross revenue calculation as spendable income.
For a small site, the questions are straightforward: can I receive payouts, what deductions apply, how much must accrue, and when will it arrive?
Will charging AI bots help or hurt SEO?
Do not treat paid crawling, search visibility and human visits as one metric. Selling access does not purchase a citation, a ranking or a reader.
Before changing access, identify the exact crawler and its purpose. Keep a record of which rules you changed. A blanket policy is a poor experiment if you cannot tell whether it affected the bots you meant to charge or another route through which people discover your work.
Track organic search clicks and relevant referral visits separately from paid requests. Then track what those people do: subscribe, return, contact you or buy something. Our guide to paid placement versus organic AI visibility covers a related mistake: counting different kinds of exposure as though they delivered the same result.
If crawl revenue rises while valuable referrals fall, you have a trade-off to investigate, not an automatic win. If referrals fluctuate, do not blame the access change without checking other explanations.
What is worth testing on a small publication?
Start with an inventory of the work that would be hard to replace. Perhaps you maintain a niche supplier database, interview people others cannot reach, or publish documented product tests. These are reasons to investigate demand, not evidence that an AI company will pay.
A rewritten announcement gives a prospective buyer plenty of substitutes. An original comparison may offer something different. The commercial question is still whether anyone needs that difference enough to purchase access.
For a first test, keep a short record:
- Before the change: participating crawler activity, organic clicks and relevant referrals over a defined period.
- The change: selected crawlers, price, date and any overlapping security rules.
- The result: confirmed paid requests, reconciled revenue and eventual payout, alongside reader activity.
- The cost: time spent producing and maintaining the material, plus time managing the experiment.
If nobody buys, find out whether the issue is access, configuration or lack of demand before commissioning more content for the bots.
Where AntiMedia fits
For a publisher or an iGaming media team, the useful work often happens before publication: checking a tool's terms, testing a reporting screen or asking someone to explain where a product page becomes confusing.
That is work you can brief. Through our own service, AntiMedia Earn, brands and media teams can propose marketing and research tasks, with deliverables and proof for review. For example, commission a source-linked pricing check or a documented product walkthrough. Editors still need to verify the findings, respect confidentiality and disclose commercial involvement where relevant.
This is not a way to manufacture positive reviews or guarantee AI citations. It is a way to commission material that can give an article something concrete to say. If your team has that kind of assignment, send us the brief.
Pay Per Crawl is worth watching because it creates a way to charge for access. For a small publisher, it becomes a business result only when a buyer pays and the economics make sense. Until then, keep publishing for the people you want to reach. A pricing setting is not an audience.
