Tools: 77
Marketing

Ahrefs

Connect your account, then chat with AI to run tools.

Ahrefs SEO over REST v3: 75 typed tools covering the official ahrefs-python v3 surface (Site Explorer, Keywords Explorer, Rank Tracker, Brand Radar, Site Audit, SERP overview, batch analysis, management, public metadata) plus four enum-driven Web Analytics tools (tabular, summary chart, breakdown charts, UTM) and raw GET/mutate helpers. Dashboard API key (Bearer) against api.ahrefs.com/v3.

Use in your AI tool

You’ll sign in to MCPBundles when your client connects.

MCP URL
https://mcp.mcpbundles.com/bundle/ahrefs

Chat in MCPBundles Studio

This server needs a sign-in. MCPBundles handles that setup once — then chat here with saved access and a log of what ran.

Managed
77 tools
Agent guide included

Opens MCPBundles Studio with this server selected. After sign-in, chat and run tools from the same thread.

Browse all tools

Live probe refreshed Sep 14, 2026 · Endpoint host mcp.mcpbundles.com

AI Skill
SKILL.md

Domain knowledge for Ahrefs — workflow patterns, data models, and gotchas for your AI agent.

Ahrefs

Ahrefs spans six research surfaces. Pick the surface that matches the question before pulling data — competitor recon, rank monitoring, keyword discovery, AI-search visibility, technical crawls, and first-party traffic are different products with different ids.

First moves

  1. Usage check — read subscription limits and current API unit consumption before large scans.
  2. List projects — resolve the numeric project id when the user names a tracked site (Rank Tracker, Web Analytics, or management reads).
  3. Site Explorer snapshot — domain-level metrics, organic keywords, backlinks, or top pages for any public target without a project id.
  4. Rank Tracker overview — tracked-keyword positions for an existing project when the user asks "where do we rank today".
  5. Keywords Explorer — ad-hoc keyword difficulty, volume, and SERP context for seeds not yet in a Rank Tracker list.
  6. Brand Radar report — AI-search mentions, citations, and share-of-voice for a configured report id.
  7. Site Audit health — crawl issues, affected pages, and link graph for a verified Site Audit project.
  8. Web Analytics — first-party visit trends for a connected analytics project (not Site Explorer traffic estimates).

Pick the right surface

  • Site Explorer — "What does Ahrefs know about this domain or URL on the public web?" Competitor research, link audits, traffic estimates, and one-off lookups. Pass any target domain or URL; no project setup required.
  • Rank Tracker — "Where does my tracked project rank for my keyword list today vs yesterday?" Ongoing monitoring only — keywords must already live in a Rank Tracker project.
  • Keywords Explorer — "Should we target this keyword?" Research volume, difficulty, parent topics, and SERP features for seeds the user has not added to Rank Tracker yet.
  • Brand Radar — "How visible are we in AI answers?" Mentions, citations, and share-of-voice across ChatGPT, Perplexity, Google AI Overview, Gemini, and similar surfaces for tracked prompts.
  • Site Audit — "What is broken on our site technically?" Crawl health, issue counts, internal links, and page-level flags after domain ownership is verified.
  • Web Analytics — "What is happening on our site right now?" First-party visit and conversion data piped into Ahrefs — not the third-party traffic model Site Explorer shows.

Project and report ids

  • Rank Tracker and Web Analytics share one numeric project_id. List projects and match the tracked URL — never guess the integer.
  • Site Audit uses a separate project namespace; a Rank Tracker project_id will not work for audit reads. Most audit work requires verified domain ownership on that project.
  • Brand Radar uses report_id (numeric, from the Brand Radar report URL). Brand Radar reads never accept a Rank Tracker project_id.
  • Keyword lists use keyword_list_id — a standalone list surface distinct from project tracked keywords.

Operational order

  1. When the user says "rankings" without context, ask whether they mean tracked project positions (Rank Tracker) or market research on a domain (Site Explorer / Keywords Explorer).
  2. Resolve project_id or report_id from a list call before drilling into project-scoped reads.
  3. For competitor domains, start with Site Explorer metrics and organic keywords; switch to Rank Tracker only when monitoring a saved keyword set.
  4. Apply tight limit values on row-priced reads; widen only when the user needs exhaustive exports.

Rank Tracker maintenance

When adding competitors or tracked keywords, payloads are object arrays, not bare strings.

  • Competitor items: { "url": "<domain or url>", "mode": "domain" | "subdomains" | "exact" | "prefix" }. Prefer subdomains for whole-brand tracking (www, app, m, etc.); domain excludes subdomains.
  • Tracked-keyword items: locations as [{ "country": "us", "language": "en" }, ...] and keywords as [{ "keyword": "...", "tags": ["..."] }, ...] (tags optional).
  • Standalone keyword-list updates accept a flat string array of keywords — different shape from project tracked-keyword writes.

Column-name traps (select)

Each report exposes its own column identifiers; the wrong name returns an error listing allowed columns. Repeated traps:

  • Site Explorer organic keywordsbest_position and best_position_url; position and bare url are invalid there.
  • Rank Tracker overviewposition and url for the tracked keyword's current rank.
  • Organic competitors — keywords_common, not common_keywords.

Web Analytics date argument

Tabular and chart reads accept to and from_ (trailing underscore on from_). Pass from_ in tool arguments; upstream receives the plain from field name.

Date formats

  • Site Explorer, Rank Tracker, Keywords Explorer, Brand Radar, Web Analytics — YYYY-MM-DD or ranges with the same format.
  • Site Audit crawl slices — YYYY-MM-DDThh:mm:ss timestamps; omit the date to read the latest stored crawl.

API units

Most analytics reads consume API units against the monthly allowance; management lists and Rank Tracker writes are typically free. Row-priced surfaces (Site Explorer, Keywords Explorer, Brand Radar, Web Analytics) bill per row returned — keep limit low while exploring.

Tools discovered (77)

Showing 25 of 77 from the live probe.

  • Ahrefs Ba Batch Analysis Post

    Pull selected SEO metrics for a batch of up to 200 targets (domains or URLs) in one call. ``targets`` is a list of objects, each with ``target`` (domain or URL), ``mode`` (``domain`` / ``subdomains`` / ``exact`` / ``prefix`` / ``url``), and optional ``protocol``. ``select`` lists the columns to return — see the column-name traps in the bundle skill: Site Explorer organic-keywords uses ``best_position`` and ``best_position_url``, organic-competitors uses ``keywords_common``. ``country`` (ISO-2) a

  • Ahrefs Br AI Responses

    Return raw AI-search engine responses (ChatGPT, Perplexity, Gemini, Google AI Overview, etc.) captured for the prompts tracked in a Brand Radar report. Each row is a single answer with the source engine, country, the prompt that triggered it, the answer text, and the citations rendered. Use this to audit *what AI assistants are actually saying* about a tracked brand, not just whether they mentioned it. Filter by ``date_from`` / ``date_to``, ``country``, ``data_source`` (the AI engine), and promp

  • Ahrefs Br Cited Domains

    Return the external domains that AI search engines are citing alongside (or instead of) the tracked brand for the report's prompts, with citation counts and share. Use this to discover the authoritative sources LLMs trust for the brand's topic — the domains to guest-post on, build links from, or compete against. Filter by date range, country, and data source (AI engine). Costs units per row.

  • Ahrefs Br Cited Pages

    Return the individual pages (URLs) that AI search engines cite in answers to the report's prompts, with citation counts. One step deeper than the cited-domains sibling — use this when you need to know exactly which articles are being surfaced (e.g. to study their format, length, or schema). Filter by date range, country, and AI engine. Costs units per row.

  • Ahrefs Br Impressions History

    Time-series of AI-search impressions (the brand was mentioned or cited in an answer) for the report's tracked prompts, bucketed by day. Use this to chart whether the brand's AI-search visibility is trending up or down. Filter by date range, country, and AI engine. Pair with the impressions-overview sibling for a single-number summary. Costs units per data point.

  • Ahrefs Br Impressions Overview

    Aggregated total of AI-search impressions for the report's tracked prompts over the given window — one number you can put on a dashboard. Use the impressions-history sibling for the daily series. Filter by date range, country, and AI engine. Cheaper than the history endpoint when you only need the total.

  • Ahrefs Br Mentions History

    Time-series of brand *mentions* (the brand name appeared in the answer text) for the report's prompts, bucketed by day. ``Mention`` differs from ``citation``: a mention is a name reference in the prose, a citation is a clickable source link. Mentions track brand awareness in AI answers; citations track link earning. Filter by date range, country, and AI engine. Costs units per data point.

  • Ahrefs Br Mentions Overview

    Aggregated total of AI-search mentions of the brand for the report's prompts over the given window. Use the mentions-history sibling for the daily series. Mentions = name appears in answer text (different from citations, which are clickable source links — see the cited-domains/cited-pages tools for those). Filter by date, country, AI engine.

  • Ahrefs Br Sov History

    Time-series of share-of-voice across AI search engines: the percentage of answers to the report's tracked prompts that mention or cite the brand vs. competing brands. Use this to track whether the brand is gaining or losing AI-search ground over time. Filter by date range, country, AI engine. Costs units per data point.

  • Ahrefs Br Sov Overview

    Aggregated share-of-voice for the brand across AI search engines over the given window — one number per engine/country, plus the competing brands' shares. Cheaper than the sov-history sibling when you only need the period total. Use to answer 'who is winning AI search for this category right now'.

  • Ahrefs Get V3

    Escape-hatch GET against any documented Ahrefs v3 endpoint when no typed tool covers it (or when a new endpoint hasn't been wired up yet). ``resource_path`` is the path under v3 such as ``site-explorer/domain-rating`` or ``keywords-explorer/overview`` (no leading slash, no scheme). ``query_params`` is an optional JSON object of query parameters; arrays are flattened to comma-separated strings per Ahrefs convention. Prefer the typed read tools when one exists — they have better parameter validati

  • Ahrefs Ke Matching Terms

    Return keywords that contain the seed term as a substring or phrase match — Ahrefs's Keywords Explorer 'Matching terms' report. Use this to expand a seed into the long tail (e.g. seed = ``mcp server`` returns ``free mcp server``, ``best mcp servers``, ``what is mcp server``). Filter by ``country`` (search-engine market), ``volume_min`` / ``volume_max``, ``kd_min`` / ``kd_max`` (Keyword Difficulty), ``intent``, SERP features. Tighter than the related-terms sibling, which uses semantic similarity.

  • Ahrefs Ke Overview

    Return Keywords Explorer overview metrics for one or many keywords: monthly search volume, Keyword Difficulty (KD), CPC, traffic potential of the top-ranking page, parent topic, search intent labels, and SERP features rendered (AI Overview, Featured Snippet, People Also Ask, etc.). Pass ``keywords`` as an array; ``country`` (ISO-2) is required and scopes the search engine. Use this as the first tool when qualifying any keyword candidate. Costs units per keyword.

  • Ahrefs Ke Related Terms

    Return keywords *semantically* related to the seed term — Ahrefs's 'Related terms' report. Casts a wider net than matching-terms (which is a substring match) — useful when you want adjacent topics, not just paraphrases of the seed. Filter by ``country``, ``volume_min`` / ``volume_max``, ``kd_min`` / ``kd_max``, ``intent``. Costs units per row.

  • Ahrefs Ke Search Suggestions

    Return Google-autocomplete-style suggestions that *start with* the seed phrase — the phrases users actually type when their query begins with this word. Lighter and different from the related-terms / matching-terms reports, which are SEO databases rather than autocomplete scrapes. Use for early-stage topic discovery and for long-tail expansion where intent is freshness-sensitive. Filter by ``country``.

  • Ahrefs Ke Volume By Country

    Return per-country search volume for a single keyword — every country Ahrefs has data for, with each country's monthly volume. Use this to decide which markets to localize content for, or to disprove a US-centric volume estimate (some terms have 10× volume outside the US). Costs units per keyword (not per country row, which makes it cheaper than calling overview once per country).

  • Ahrefs Ke Volume History

    Return monthly search-volume time series for a single keyword over the requested history window. Use this to detect seasonality (e.g. ``halloween costumes``), trend (``mcp server`` is up 5×), or one-off spikes. ``country`` (ISO-2) scopes the search engine. Pair with the keyword-overview tool when you want both 'today's volume' and 'how today compares to last year'. Costs units per keyword.

  • Ahrefs Mg Brand Radar Prompts

    List the AI-search prompts currently tracked in a Brand Radar report, with each prompt's text, country scope, and id. ``report_id`` (numeric, from the Brand Radar report URL) selects the report. Use before adding/deleting prompts so you can show the user what's already tracked. Free per Ahrefs management quota.

  • Ahrefs Mg Brand Radar Prompts Delete Put

    Remove tracked AI-search prompts from a Brand Radar report. ``report_id`` (numeric) selects the report; ``prompts`` is the list of prompt strings to drop, and the optional ``countries`` filter scopes the deletion to a subset of country codes (omit to delete the prompts across every country they were tracked in). Use this when pruning superseded queries from a brand-monitoring report.

  • Ahrefs Mg Brand Radar Prompts Post

    Add custom AI-search prompts to an existing Brand Radar report so Ahrefs starts tracking the report's brand presence in answers to those prompts across the given countries. ``report_id`` (numeric, from the Brand Radar report URL) routes the call; ``countries`` is a list of ISO-3166-1 alpha-2 codes and ``prompts`` is a list of UTF-8 strings up to 300 chars each. Use this when the user asks to monitor specific search queries in AI Overview / SGE responses. Free per Ahrefs management quota.

  • Ahrefs Mg Keyword List Keywords

    Return the keywords stored in a saved Keywords Explorer keyword list, with the list's metadata. ``keyword_list_id`` (numeric) selects the list. Use this to inspect what's in the list before mutating it, or to feed the keyword list into the keywords-explorer overview tool for batch metrics. This is the standalone keyword-list surface — different from Rank Tracker's tracked keywords (project-keywords). Free per Ahrefs management quota.

  • Ahrefs Mg Keyword List Keywords Delete Put

    Remove keywords from a saved Keywords Explorer keyword list. ``keyword_list_id`` (numeric) selects the list and ``keywords`` is the flat array of keyword strings to drop. Idempotent: keywords that aren't in the list are silently skipped. Free per Ahrefs management quota.

  • Ahrefs Mg Keyword List Keywords Put

    Append keywords to a saved Keywords Explorer keyword list. ``keyword_list_id`` (numeric) selects the list — discover it via the matching read tool. ``keywords`` is a flat array of keyword strings (this is the standalone keyword-list surface; Rank Tracker tracked-keyword writes use a different object shape). Idempotent: adding a keyword that already exists is a no-op. Free per Ahrefs management quota.

  • Ahrefs Mg Locations

    Return the list of locations Ahrefs supports for a given country: country/region/city labels with their numeric ``location_id`` and ``language_code``. ``country_code`` (ISO-3166-1 alpha-2) is required; ``us_state`` further narrows to a single US state. Use before calling rank-tracker or keywords-explorer endpoints that take a ``location_id``, or before adding tracked keywords with locale-specific tracking. Free per Ahrefs management quota.

  • Ahrefs Mg Project Competitors

    List the competitors currently tracked for a Rank Tracker project: each row has ``url`` and ``mode`` (``domain`` / ``subdomains`` / ``exact`` / ``prefix``). ``project_id`` (numeric) selects the project — discover it via the projects list tool by matching the tracked URL. Use before adding or deleting competitors so the user can see what's already tracked. Free per Ahrefs management quota.

Frequently Asked Questions

What is the Ahrefs MCP server?

Ahrefs SEO over REST v3: 75 typed tools covering the official ahrefs-python v3 surface (Site Explorer, Keywords Explorer, Rank Tracker, Brand Radar, Site Audit, SERP overview, batch analysis, management, public metadata) plus four enum-driven Web Analytics tools (tabular, summary chart, breakdown charts, UTM) and raw GET/mutate helpers. Dashboard API key (Bearer) against api.ahrefs.com/v3. It provides 77 tools that AI agents can use through the Model Context Protocol (MCP).

How do I connect Ahrefs to my AI agent?

Add the MCPBundles server URL to your MCP client configuration (Claude Desktop, Cursor, VS Code, etc.). The URL format is: https://mcp.mcpbundles.com/bundle/ahrefs. Authentication is handled automatically.

How many tools does Ahrefs provide?

Ahrefs provides 77 tools that can be called by AI agents, along with a SKILL.md that gives your AI agent domain knowledge about when and how to use them.

What authentication does Ahrefs require?

Ahrefs uses API Key. Ahrefs requires credentials. Connect via MCPBundles and authentication is handled automatically.

Setup Instructions

Connect Ahrefs to any MCP client in minutes

Ready to use Ahrefs?

Sign in to connect your credentials and start running tools from the chat.