Once you have a list of cited domains, a few API calls turn it into a comparison table. Here is the method, the calls and what the numbers will not tell you.
AlgoLink research team ·
Four calls, about a cent in total
01
Collect the domains
From your experiment log, take each distinct cited domain. Strip paths and www. Fifty domains is typical for a topic, and the source claims explain why you should repeat the list monthly.
02
Score them in bulk
bulk_dr takes up to 2,000 domains per call and costs one credit, which is $0.0001. You get a 0 to 100 DR for each.
03
Open the interesting ones
For three or four, call refdomains (2 credits) and read who links. A call is billed once whatever the row count, up to 1,000 rows.
04
Look for overlap
intersect accepts 2 to 10 domains at 2 credits and returns sites linking to several of them. Useful for outreach ideas.
The calls
Replace the key with your own. The domains below are placeholders.
bash
# 1 credit: DR for many domains
curl "https://algolink.net/api.php?action=bulk_dr&domains=example-a.com,example-b.org,example-c.net&key=sk_live_..."
# 2 credits: referring domains for one of them
curl "https://algolink.net/api.php?action=refdomains&q=example-a.com&per=100&key=sk_live_..."
# free, no key: quick overview
curl "https://algolink.net/api.php?action=overview&q=example-a.com"
What the result might look like (hypothetical)
Invented numbers to show the shape of the output, not findings.
Cited domain
DR
Reading
example-a.com
91
Large, broadly linked publisher
example-b.org
68
Mid-sized reference site
example-c.net
34
Niche site, few links, still cited
your-site.com
22
Below most of the set, not absent from it
Reading the table without fooling yourself
DR here is our own metric: round(100 times log(ref_domains + 1) divided by log(20,000,000)). It rises slowly. Going from 10 to 20 means far more links than going from 60 to 70.
Spread first. Average later. If a topic's cited set runs from DR 20 to DR 90, links are plainly not the entry ticket. A tight cluster at the top would be more suggestive, though still not causal.
Mind the lag. Our index is refreshed roughly monthly, so a link from last week may not show. Ahrefs and Semrush crawl more often and track keywords, which we do not. Pick them when freshness is the point.
One more use. The comparison also feeds a second question. Which sites link to several cited domains and not to you? Those are candidates for outreach, and the myth check covers the cheaper mention route.
Cost check. Fifty domains scored, four referring-domain pulls and one intersect come to roughly 1 + 8 + 2 = 11 credits. That is about a tenth of a cent, so the limit on this method is your attention, not your budget.
Two traps to avoid
Do not compare DR across very different topics, because a cooking set and a cybersecurity set have different natural ranges. And do not treat missing links as proof of absence: our monthly refresh means recent links may not be listed yet.
Questions about the pull
How do I search backlinks for free?
The domain, overview, stats and suggest actions need no key and cost nothing, so you can look up any domain's headline numbers. Row-level lists such as referring domains cost 2 credits each, covered many times over by the $100 sign-up credit.
Does a higher DR mean an AI tool will cite me?
No evidence supports that. DR summarises how many distinct sites link to a domain. Cited sets we would expect to include low-DR sites, so use DR to describe the neighbourhood, never to predict a citation.
How many domains can I score at once?
Up to 2,000 per bulk_dr call, sent as a GET parameter or in a POST body. It costs one credit regardless of count, so batching is almost always cheaper than looping.