Introduction

The RankParse API is a REST API over HTTPS. It takes query parameters or JSON bodies and always answers with JSON (apart from the two legacy redirects and unknown routes). Every page in this reference documents one operation: its parameters, its response fields, what it costs and the errors it returns.

Base URLCopy Anchor Link

https://api.rankparse.com/v1

Every path in this reference is relative to that base, so GET /v1/backlinks is https://api.rankparse.com/v1/backlinks. Requests must use HTTPS. The API allows cross-origin requests, but a key must never be shipped in browser code (see Authentication).

AuthenticationCopy Anchor Link

Send your API key in the X-API-Key header on every request. Keys start with rp_ and are created under Settings, Developers, API keys in the dashboard.

curl "https://api.rankparse.com/v1/domain-authority?domain=github.com" \
  -H "X-API-Key: rp_your_key"

A missing, unknown or revoked key returns 401 invalid_api_key. A key in the query string is not read. The public GET /v1/pricing and the free tools are the only routes that need no key.

RequestsCopy Anchor Link

• GET operations take query parameters. Domains are bare hostnames such as stripe.com, never URLs: https://stripe.com returns 400 invalid_domain. URL parameters (url=) take a full public http or https URL.
• POST and PATCH operations take a JSON body. Send Content-Type: application/json.
• A few write operations keep their target in the query string, for example POST /v1/indexnow/key?domain=, POST /v1/site-scan/recheck?site_id=, POST /v1/site-audit/run?site_id=, PATCH /v1/site-context?site_id= and PATCH /v1/rank-tracking/alert-settings?site_id=. Each operation page says where every parameter goes.
• site_id is the id of one of your websites. List them with GET /v1/websites or GET /v1/rank-tracking/sites.

ResponsesCopy Anchor Link

Billable endpoints wrap their result in one envelope:

{
  "domain": "stripe.com",
  "data": [
    {
      "from_domain": "example-blog.com",
      "from_url": "https://example-blog.com/payments",
      "to_url": "https://stripe.com/",
      "anchor_text": "Stripe",
      "rel": null,
      "link_type": "hyperlink",
      "crawled_at": "2026-02-11"
    }
  ],
  "total": 18342,
  "limit": 1,
  "credits_used": 2,
  "credits_remaining": 998
}

Fields that do not apply are left out rather than sent as null. Free routes (account, rank tracking, Search Console, Google Ads, issues, indexing, AI visibility, site audits, project context, the approval inbox and outreach) have their own shapes, documented on their pages. The details, with examples, are in Response format.

PaginationCopy Anchor Link

• limit defaults to 100. On the link graph endpoints it is clamped to 1 to 1000 (similar-domains stops at 100, competitor-gap defaults to 50). A non-numeric value falls back to the default.
• The link graph endpoints return the top limit rows in their own order and take no offset.
• A few routes page with limit and offset: /v1/usage, /v1/indexing and the rank tracking lists. Their pages give the exact limits.
• A domain with no data returns 200 with data: [] and is billed as usual. RankParse never answers 404 for "no data".

ErrorsCopy Anchor Link

Errors return a non-2xx status and a JSON body with a stable code. Branch on code, not on the message text.

{
  "error": "Payment Required",
  "code": "insufficient_credits",
  "message": "Add credits at https://rankparse.com/pricing"
}

Every code, with what to do about it, is in Errors.

Rate limitCopy Anchor Link

Each API key may make 60 requests per minute, counted over a sliding 60 second window. Going over returns 429 rate_limited with no Retry-After header. Two keys do not share a counter. The account routes (/v1/credits, /v1/me, /v1/keys, /v1/usage, /v1/pricing) and the connection status routes are not counted. PageSpeed and Google Ads also have daily caps. See Rate limiting.

CreditsCopy Anchor Link

• Each billable call is charged in one atomic step, before the work runs. The cost is on every operation page and in the table on Credits.
• Monthly plan credits (Pro and Agency) are spent first, then purchased credits. credits_remaining is the sum.
• A call that fails on the server side (a data store timeout, a failed live fetch, a failed IndexNow submission) is refunded automatically, and its credits_used reads 0.
• Every data lookup and live page endpoint validates its parameters before charging, so a 400 for a missing or invalid parameter costs nothing.

Caching and retriesCopy Anchor Link

• Link graph answers are cached for 30 days per crawl release, endpoint and parameters. A cached answer is billed exactly like a fresh one and is the same data. Nothing in the response says whether it came from the cache. page-performance results are cached for 24 hours per URL and strategy.
• There is no Idempotency-Key header. Retrying a billed call that succeeded charges it again; retrying one that failed with a refunded error is free.
• Some writes are safe to repeat: POST /v1/indexnow/key returns the same key for a host, a proposal for the same action returns the existing proposal, and POST /v1/activity takes an optional client_request_id that makes a repeat return the first note.
• A 503 query_timeout or upstream_error carries retry: true, retry_after: 3 and a Retry-After: 3 header. deterministic_timeout carries retry: false: change the request instead of retrying it.

ResourcesCopy Anchor Link

Link graphCopy Anchor Link

Competitive analysisCopy Anchor Link

Live page analysisCopy Anchor Link

PlatformsCopy Anchor Link

Your websitesCopy Anchor Link

Agent actionsCopy Anchor Link

AccountCopy Anchor Link