Reference
Find a website's backlinks: every call, priced and rated
Each action answers one valuation question and bills once per call, however many rows come back. Read the table, then use the notes on how each answer can mislead.
Calls in one table
| Action | Credits | Answers | Misleads when |
|---|---|---|---|
| domain, overview, stats, suggest | 0 | Is the domain known, and what is its DR | You treat DR as a quality verdict |
| bulk_dr | 1 | DR for up to 2,000 domains | Domains are new or recently moved |
| compare | 1 | Up to 20 domains side by side | The domains differ in age by years |
| top, dr_distribution, tld_breakdown | 1 | Shape of a profile: strongest sources, DR spread, TLD mix | You read a single bar without the total |
| refdomains, backlinks | 2 | Who links, and the individual links | You count rows as links from separate sites |
| anchors, anchor_cloud | 2 | The wording used by linkers | A brand name hides a paid campaign |
| toppages | 2 | Which URLs on the target hold the links | The page moved since the last refresh |
| intersect | 2 | Sources linking to 2-10 domains at once | Your rivals have few links to intersect |
| search, keyword | 3 | Full-text over anchors, titles, URLs; who is linked with an exact anchor | The phrase is generic |
Notes on reading the answers
Each call returns JSON with an ok flag, and authenticated calls add a _credits object with the cost and your remaining balance. Check the key names in a live response before wiring them into code.
refdomains
Sort with
sortandorder, page withpageandper(10-5000). One call atper=1000costs 2 credits, so a full thousand-row pull is $0.0002.backlinks
Use it when you need the exact page that links, for outreach or to confirm a placement. Do not use it to count sites; refdomains is cleaner.
anchors
Read the top five rows first. If one commercial phrase leads, the profile was probably built to order.
toppages
The most useful call before a redirect. It names the URLs whose loss would hurt, which is the subject of the migration post-mortem.
intersect
Hand it two to ten domains and it names the sources that link to several. Those are usually editorial sites, which makes them the better ones to approach.
bulk_dr
The cheapest call by far. One credit ranks a whole offer list, so run it before anything else that costs 2.
Paging a large profile
Ask for the biggest pages once, stop when the next page adds nothing.
curl "https://linkmidasai.com/api.php?action=refdomains&q=example.org&per=1000&page=1&key=sk_live_..."The index is our own crawler's view, refreshed roughly once a month. A missing row means not seen yet, not never linked. HTTP 401 means a bad key, and HTTP 402 means the balance is spent.
Questions the reference does not settle
Which call should I run first when I need to find a website's backlinks?
Start with the free overview, then run refdomains for 2 credits. Overview tells you whether the domain is worth a paid pull, and refdomains gives you the list of distinct sites. Add anchors and toppages only when the first two look interesting; the recipe runs them in that order.
Does the same link count twice if it appears on many pages?
The backlinks call lists individual links, so yes, one site can fill many rows. The refdomains call collapses them to one row per linking domain, which is why it is the better basis for a price.
Can I get more than 1,000 rows?
Yes, by paging. Each call is billed once for up to 1,000 rows, and deep pagination is capped at 50,000 rows. For bigger profiles, slice by sort order and keep the top of each list. The scoring model needs only the top rows anyway.