SE Ranking API and MCP: Credits, Limits and Setup

What the API gives you, in short

SE Ranking's API is included in every plan and also sold on its own ^[1]. One key covers two jobs: raw SEO data for research, dashboards and products, and the management of the projects inside your account. Core includes 25,000 Data API credits a month and Growth 100,000, and every research call is priced in those credits, from 1 credit per backlink to 7,500 for an AI search leaderboard. On top of the API sits a hosted MCP server that puts the same data inside Claude, ChatGPT and other assistants, paid from the same credits. This guide draws on SE Ranking's documentation and help center as of 12 October 2026.

The practical summary is that the endpoint you call decides the bill more than the plan does. The Data API allows 10 requests per second and the Project API 5, keys cannot be scoped, and plan credits do not roll over. It still rewards planning.

Two APIs behind one key

SE Ranking splits its API in two. The Data API serves raw research data, while the Project API manages the projects inside your account, and both are served from api.seranking.com. In the vendor's words, "A single API key authenticates every endpoint across the Data API and the Project API." There is no second credential to manage. Two smaller APIs sit alongside the main pair: a Local Marketing API for locations, local rankings, local audits, Google Business Profile metrics, listings and reviews, and an SE Visible API, marked beta, for projects, brands, aliases, topics and prompts.

The Data API: research data

The Data API covers keyword research, backlinks, domain analysis, website audits and AI search, and the documentation groups its endpoints into those same five families. Typical calls return keyword metrics, backlink lists, a domain's traffic by region, audit issues and a brand's AI search presence. Backlink endpoints cover summaries, lists, new and lost links, anchors, referring domains and authority, while domain endpoints cover traffic by region, rankings, pages, subdomains, paid ads and keyword gaps. It is the half of the API that spends credits.

The Project API: your account, scripted

The Project API manages projects: creating sites, adding keywords and search engines, running position checks, and reading rankings, competitors, audits and backlink monitoring. Its endpoint list also covers keyword groups, URL tags, Search Console data, sub-accounts, the marketing plan, disavow lists and the AI Results Tracker. Almost anything you click in a project can be scripted. It draws on your plan's limits, not on credits: anything that counts against a limit in the interface counts in the API too. For agencies that mainly want rankings in their own reports, it is the cheaper half.

How credits work, and what calls cost

The charging rules

Successful Data API calls are charged either per record returned or as a flat number per request. Failed requests, the 4xx and 5xx errors, do not consume credits. Plan credits renew each cycle and do not roll over. Unused credits are lost. Checking the balance through the Account System API costs 0 credits. One catch applies. The subscription endpoint does not include Wallet credits, which appear only in the API Dashboard.

The price list that matters

The spread between endpoints is wide, and it is the single most important thing to read before writing code ^[2]:

What Core's credits actually buy

Core's 25,000 monthly credits pay for three AI leaderboards, or 31 AI overviews, or 25,000 backlink records, or 250 bulk keyword requests. Growth's 100,000 credits buy four times as much. The lesson is plain. Check the cost of each endpoint before writing a loop that runs every night, then test small, because a nightly job that calls the AI search endpoints for a handful of brands can empty the same allowance that would cover 25,000 backlink records.

The trial, add-ons, the Wallet and standalone plans

The free trial

The trial includes 100,000 API credits across every endpoint, and they expire with the 14-day trial. The Data API is limited to 1 request per second during the trial, a cap that can be raised on request. MCP is a grey area. The help center lists MCP access as not available on the trial, while the API pricing page includes it. Do not plan an evaluation around it.

Three ways to buy more

There are three routes to more credits ^[3]. The API add-on extends a Core, Growth or Enterprise plan, on annual billing only. The standalone API needs no platform subscription and starts at US$179 a month, also annual only, advertised at up to 57% off Wallet rates with volume discounts. Accounts on a standalone Data API license get a streamlined interface that goes straight to the API Dashboard, with only the API section, the Wallet and MCP in the menu, so developers who never need the SEO platform are not paying for it. Enterprise starts at 120M+ credits a year, priced as low as $30 per 1M credits, with custom rate limits and custom integrations.

The Wallet

The Wallet sells pay-as-you-go credits from US$50 for 250,000, and purchased credits never expire. Plan and add-on credits are used first; the Wallet is drawn down only after them. A top-up can even reactivate Data API access after a plan or trial has expired. At US$50 per 250,000 credits, an AI leaderboard costs about US$1.50. Enterprise's best rate is more than six times cheaper per credit than the Wallet's US$0.20 per 1,000.

Two rules trip people up. Wallet top-ups are available on Core, Growth and other paid plans, from the main account only, so client accounts in an agency workspace and sub-accounts cannot top up. And the general SE Ranking balance, which pays for tools like Keyword Grouper, does not reactivate Data API access on an account whose trial has expired; only a plan with API credits or the API Wallet does that. Auto-recharge and low-balance alerts are still marked coming soon, so monitor the balance yourself for now.

Where SE Ranking's own pages disagree

The add-on numbers do not line up across SE Ranking's pages. The pricing page sells the US$45 add-on with 3M credits, plus bigger packages at US$149 and US$555, while the API pricing page sells the same US$45 add-on with 250K credits a year. Standalone access shows the same split: the documentation says it starts at 24M credits a year, against 1M credits a year on the API pricing page. These are not rounding differences. Confirm the credit count in writing before you buy an add-on or a standalone plan.

Getting a key, limits and errors

From key to first response

Open the API Dashboard, click Create API Key and copy it. A standalone Data API account goes straight to the API Dashboard on sign-in. SE Ranking publishes a Postman collection for first requests: import it, fork the environment, paste the key under secret_key_data, then call Account and system, subscription. A response with your subscription details confirms access. From there, the same key reaches every endpoint.

Rate limits and bulk calls

The Data API allows 10 requests per second and the Project API 5, per API key. A 429 error means too many requests. Batch work should pause between calls. Bulk endpoints help on both fronts: keyword metrics, for example, takes up to 5,000 terms in one request, which saves credits and rate limit at the same time. Plan for that early.

Errors that look like billing problems

Three status codes cause most of the confusion. A 402 Insufficient Funds error right after purchase usually means the key is not yet linked to credits, and waiting a few minutes fixes it. A 401 is an authentication failure, not a credit problem. A 403 means the account lacks Data API access, even if the Wallet shows credits. Check entitlement first.

Keys you cannot restrict

"API keys cannot currently be scoped or restricted" ^[4]. There are no IP allowlists or per-endpoint permissions. The last remaining Data API key cannot be deleted until you create a replacement. The defense is procedural. Give each integration its own key, and rotate it when people leave. No exceptions.

MCP: the API inside your assistant

Hosted, and paid from your credits

SE Ranking's MCP server is hosted by SE Ranking and needs only a URL and authorization, so there is nothing to install. It uses OAuth 2.1 at api.seranking.com/mcp, and SE Ranking says it exposes 180+ SEO and GEO tools. MCP access is included in every SE Ranking subscription, but the data it pulls is paid from the plan's API credits. You can also run MCP on a pay-as-you-go Wallet without a subscription. For most teams it is the simpler route than code, because the assistant handles the calls and the parsing.

Connecting Claude, ChatGPT and other clients

In Claude, add the official SE Ranking connector under Settings and Connectors; in ChatGPT, search for SE Ranking under Apps. Any other MCP client can add a custom connector with the URL and authorize with OAuth or an API key. The MCP page lists Codex, Cursor, VS Code, v0 and Windsurf among supported clients, and the help center also explains connecting the server to Gemini. A test prompt, such as asking for a domain's organic stats, confirms the link.

Skills and ready-made prompts

Skills are reusable instruction files that tell an assistant which MCP tools to call and what deliverable to produce. SE Ranking publishes skills for content briefs, AI search share of voice, technical audits, backlink gaps and paid search competition, installed from its marketplace in Claude Desktop or Claude Code. The MCP prompt library adds prompts for keyword ideas, domain overviews by region, keyword movements, paid ads, backlink gaps and Local Marketing reviews. Copy them and replace the placeholders.

Watching what the assistant spends

Every MCP answer can trigger several API calls. A request for an AI search leaderboard alone costs 7,500 credits, close to a third of Core's month. The same risk grows with agents: the help center describes agents that decide which keyword, backlink or domain data to fetch next rather than following a fixed script. Agents spend fast. Run heavy MCP work on Growth or with Wallet credits, and check the API Dashboard balance after the first week.

What teams build with it

Rank tracking and audits

The rank tracker API returns daily positions, top-100 results, ranking history, 35+ search features and landing pages by engine, device and location, and it can create projects, add keywords and trigger a recheck. Agencies use it to send ranking drops to Slack the same day. The Website Audit API runs the same crawler with 115+ checks, asynchronously: create the audit, poll its status, then pull the report. It returns a 0 to 100 health score, issues by severity and the pages affected. Pipe the issues into a ticket queue after each crawl.

AI search data and products

The AI Search API returns a domain's presence by engine, a share-of-voice leaderboard against up to 10 competitors, and the prompts where a brand appears. The help center pitches it for client portals and SaaS products, and budgets should allow for the 7,500-credit leaderboard calls. SaaS teams can also embed SE Ranking data instead of building their own crawlers, under the commercial use rules below.

Reporting and content without code

A native n8n node connects Data API endpoints to no-code workflows, Make has separate setups for the Data API and the Project API, and Looker Studio connectors feed rankings and AI data into dashboards. These cover most reporting automation without code. The help center adds guides for AI-powered content pipelines that pull keyword and SERP data into briefs, and for pairing MCP with GA4 data in one assistant session. Each call still uses API credits.

Two worked budgets

An agency dashboard

An agency with 20 clients pulls daily rankings through the Project API, which costs plan limits rather than credits, and one monthly AI leaderboard per client, which comes to 150,000 credits. That exceeds Growth's 100,000, so the agency needs Wallet credits, about US$10 more each month at the Wallet rate. If the same agency ran each leaderboard once a quarter instead, its monthly AI spend would fall to about 50,000 credits, well inside Growth's allowance. The choice of endpoint, rather than the plan, decides the bill.

A link-building shortlist

SE Ranking's link-building quickstart prices each step: filtering domains by authority costs 10 credits per domain, a backlink summary 100 credits per domain, and top backlinks 1 credit each. Vetting 100 outreach targets with both checks costs 11,000 credits, under half of Core's monthly allowance. For one-off research jobs like this, the Wallet is often the better fit than a bigger plan, since its credits never expire.

Commercial use: who needs a contract

API data may be used for commercial purposes, subject to SE Ranking's Terms of Service. The one firm exception is direct competitors, who must get SE Ranking's explicit consent before using or distributing the data. In practice, agencies reselling reports are fine, while tools competing with SE Ranking need a contract. For a product built on the data, Enterprise terms are the realistic route. Read the terms first.

How it compares, and which setup to choose

Semrush's Standard API is an add-on to its top tier and starts with zero units until you buy them, and its MCP comes with Starter and Pro+. Ahrefs lists API and MCP access from its Lite plan. SE Ranking includes both on every plan. For a two-person agency that wants rankings in a client dashboard and an assistant that answers keyword questions, the US$129 Core plan covers both on day one, while Semrush's MCP units start with its US$199 Starter plan. Our SE Ranking pricing guide covers the plans themselves.

The setup follows the workload. For dashboards and light automation, the credits in Core or Growth are enough. For one-off research jobs, use the Wallet. For products that call the API all day, price the standalone plan or Enterprise, which offers custom rate limits and volume pricing. If the platform itself is still undecided, our SE Ranking review weighs the rest of the product.

The bottom line

SE Ranking's API is one of the most accessible in SEO: included in every plan, one key for both APIs, clear errors, a hosted MCP server and ready Claude skills. The work is in the budgeting, not the setup. Read the endpoint costs before automating, because a few AI leaderboard calls can use up a month of Core. Keep keys separate per integration, since none can be restricted. And confirm add-on credit counts in writing, because SE Ranking's own pages disagree.