how to get backlinks for free in 2026

backlinks remain the currency of seo. the problem: ahrefs charges $129/mo, semrush starts at $140/mo, and even “budget” tools sit at $30-50/mo. for freelancers, startups, and small businesses doing one-off audits, that math doesn't work.

the good news is you can survey a backlink profile for free using a stack of legitimate sources. this guide covers what's actually free, what each tool shows, and when you'll hit the wall and need to pay.

what backlinks actually are (and the metric that matters)

before tools - the definition. a backlink is one hyperlink from a page on another site to your site. a referring domain is the unique website that contains those backlinks; it can give you one backlink or ten thousand and still counts as one referring domain.

google weighs referring domain count more heavily than raw backlink count. if a single site links to you 100 times, that's one editorial vote made 100 ways. if 100 different sites link to you once each, that's 100 distinct votes - and the second profile ranks higher every time. the 2024 google search leak confirmed numUniqueLinkingDomains is a real attribute on documents. focus there. for the full breakdown, see what counts as a referring domain.

the free options, side by side

five free or freemium ways to look at a backlink profile. they show different slices of the truth, which is why most practitioners combine them.

the key thing to internalise: gsc and bing webmaster only show your own verified properties. you cannot use them to look up a competitor. for competitor profiles, you either use crawlgraph's 5-result free preview or pay once for full access.

method 1: google search console

if you own the domain, gsc is the most authoritative source - it's google's own view of what links into your site. open the property, click links, and you get top linking sites, top linked pages, and top anchor text. exports are csv and capped at roughly 1,000 sample rows per report.

caveats:

• only works for domains you've verified
• the sample is not exhaustive - google deliberately truncates
• nofollow vs dofollow attribute breakdown isn't shown

method 2: bing webmaster tools

bing's equivalent is more generous on row counts (up to ~10k per report) and often surfaces links google doesn't - different crawler, different priorities. same restriction: only your verified properties.

method 3: crawlgraph free tier

for any domain on the public web - yours, a competitor's, a prospect's - the crawlgraph homepage returns the top 5 backlinks per domain with no signup. that's enough to:

• verify a domain has a non-trivial profile before bidding on a project
• spot-check a competitor's strongest referring domains
• see whether two competitors share an outsized number of linkers

what the free tier doesn't do: full export, the 6th-through-100,000th backlink, or api access. those are on the lifetime tier.

method 4: google search operators

google's own search reveals link opportunities if you know the syntax. these operators are entirely free.

find guest posting opportunities

"guest post" "write for us" your-keyword
"submit a guest post" your-keyword
"guest article" "your-niche"

discover resource page targets

your-keyword "resource page"
your-keyword "useful resources"
"link to us" your-keyword

find journalist source queries

"journalist query" your-expertise-area
"source" "your-keyword"
"reporter" "your-keyword"

google's old link: operator is deprecated and unreliable. don't rely on it for competitor analysis - use a graph tool instead.

method 5: manual research

reverse-engineer competitors

1. identify 3-5 successful competitors
2. run each through crawlgraph's free tier to confirm a profile exists
3. for the ones worth deeper digging, use the lifetime tier or check their gsc-disclosed partnerships in their own content
4. prioritise domains that link to multiple competitors - they're open to your category
5. create content worthy of similar placement and pitch

monitor unlinked brand mentions

set google alerts for:

• your brand name and variations
• product names
• founder/team member names

many mentions occur without a link. a polite email asking for the link conversion often succeeds because the publisher already considers you reference-worthy.

where the data actually comes from

ahrefs, semrush, and moz run their own continuous crawls. crawlgraph runs on common crawl - a non-profit that has crawled the public web every 1-2 months since 2008 and publishes the result openly. the latest release as of writing is cc-main-2026-jan-feb-mar: 5.9 billion pages, 45.5 million hosts.

the practical consequence: data freshness on crawlgraph follows common crawl's release schedule - about 4-8 weeks after a crawl ships, the corresponding hyperlink graph lands. a link built today won't show up for the next two-to-three months. for backlink work older than a quarter, the cadence is rarely the bottleneck. for fast-moving link-building campaigns where you need 15-minute granularity, common crawl isn't the right data source - that's where ahrefs earns its $129/mo.

the upside of running on open data: every number we report can be reproduced by anyone willing to point duckdb at the same parquet files. for the full picture, see common crawl, explained for seos and the release schedule.

when the free stack runs out

the free stack works for: solo audits of your own site, single competitor spot-checks, and content-strategy research. it stops working when you need:

• full backlink lists for sites you don't own (gsc/bwt block this)
• csv/json exports for client reports
• programmatic access via api
• more than 5 results per domain across many domains

at that point you have two paths. ahrefs/semrush at $129-$140/mo recurring, or crawlgraph lifetime at $99 once. the lifetime tier covers up to 100,000 rows per backlinks query and ships a public api with 1,000 /api/v1/backlinks calls/mo and 50 /api/v1/gap-analysis calls/mo. details on the pricing block and api docs.

step one - submit the job:

curl -X POST https://crawlgraph.com/api/v1/gap-analysis \
  -H "Authorization: Bearer cg_live_…" \
  -H "Content-Type: application/json" \
  -d '{
    "my_domain": "yoursite.com",
    "competitor_domains": ["rival-a.com", "rival-b.com"]
  }'

the response gives you a job id and a poll url:

{
  "job_id": "gap_8c2f…",
  "status": "queued",
  "poll_url": "/api/v1/gap-analysis/gap_8c2f…"
}

step two - poll until it's done (typical jobs finish in seconds):

curl -H "Authorization: Bearer cg_live_…" \
  https://crawlgraph.com/api/v1/gap-analysis/gap_8c2f…

and the completed result, with rows you can pipe into a sheet:

{
  "job_id": "gap_8c2f…",
  "status": "completed",
  "result": {
    "total_gaps": 1284,
    "gaps": [
      { "linking_domain": "rare-site.com",      "found_on": ["rival-a.com"] },
      { "linking_domain": "industry-blog.io",   "found_on": ["rival-a.com", "rival-b.com"] }
    ]
  }
}

full reference, including pagination and error shapes, lives at /docs/api.

common mistakes to avoid

• chasing quantity over quality. a hundred links from spammy directories hurt more than they help. relevance and authority beat raw count.
• ignoring relevance signals. links from unrelated industries carry minimal weight. prioritise sites where your content genuinely belongs.
• flat anchor profiles. a natural profile has diverse anchors and a mix of dofollow/nofollow. all-exact-match looks engineered.
• expecting same-week results. common crawl publishes monthly and crawlgraph re-indexes on a quarterly composite - yesterday's link won't show up immediately.

faq

what's the difference between a backlink and a referring domain?

a backlink is one hyperlink. a referring domain is the unique website hosting those links. one site can give you a thousand backlinks and still count as a single referring domain. google weighs referring domain count more heavily than raw backlinks because each one represents a distinct editorial vote. full explanation: referring domain vs backlink.

how long does google take to count a new backlink?

three phases - crawl (4-7 days), index (1-2 weeks), ranking impact (4-12 weeks). almost everyone conflates them into one fuzzy “a few months” answer; treating them separately is what turns “why isn't this working” into “phase 2 broke and here's the fix.” full breakdown including the 2024 leak references: how long for google to count a backlink.

how often does common crawl release new data?

every 1-2 months. each release covers a few months of crawling. the corresponding hyperlink graph (used for backlink research) follows ~6 weeks after the warc files. crawlgraph reindexes within days of every graph release. the calendar lives at the common crawl release schedule.

are backlinks still a google ranking factor in 2026?

yes. the 2024 search leak named several link-related attributes (siteAuthority, numUniqueLinkingDomains, linkDecayBoost) that confirmed what practitioners had long suspected: backlinks remain one of the load-bearing inputs to ranking. the weight has shifted somewhat as google leans on click signals (navBoost) and entity-graph features, but the link graph still does most of the work for off-site ranking signals.

when do i actually need to pay for a backlink tool?

when you need full lists for sites you don't own (gsc/bwt block this), csv/json exports for client work, programmatic api access, or more than 5 results per domain across many domains. for one-off audits, the free stack above plus crawlgraph's 5-result preview covers ~90% of cases.

conclusion

finding backlinks for free in 2026 is entirely possible if you accept the constraints. gsc and bing webmaster show your own properties in detail. crawlgraph's free tier shows the top 5 backlinks per domain for any site. google operators and manual research fill the gap for prospecting.

when free runs out and you need full lists, exports, or api access, $99 once on the crawlgraph lifetime tier beats four years of ahrefs at $129/mo. start with a domain on the homepage, see what the index has, and decide from there. for methodology background, read common crawl, explained for seos.