Ahrefs Brand Radar API: What the January 2026 Update Actually Added

Ahrefs’ January 2026 release added API access to cited-page and cited-domain data from AI-generated responses, giving reporting teams a way to move beyond screenshots and manual exports. This Ahrefs Brand Radar API guide explains the concrete reporting payoff: identify which sources appear in AI answers, compare their frequency, and separate that evidence from broader claims the January update did not document.

We are deliberately keeping the scope narrow. January 2026 brought Brand Radar API additions and Page Types in Site Explorer backlink reports; it did not, based on the January release material cited here, establish every later or adjacent AI-visibility capability sometimes discussed alongside Brand Radar.

What Ahrefs released in January 2026

Ahrefs’ January 2026 materials describe new Brand Radar API coverage for cited pages, cited domains, and overview data related to AI Share of Voice distribution and history. The developer changelog dated January 2, 2026 also documents the release of endpoints concerning domains cited in AI responses.

For a brand, agency, or in-house SEO team, the most practical change is automation. Instead of reviewing a product interface and manually recording which sources appeared, a team can use API data in its own dashboard, spreadsheet process, warehouse, or client-reporting workflow.

The release supports a small but useful set of source-level questions:

• Which domains appear most often as cited sources in tracked AI-generated responses?
• Which individual pages from those domains appear?
• How often does a cited page or domain occur in the available data?
• What estimated monthly search demand is associated with the relevant mentions?
• How has the reported Share of Voice distribution changed over time?

Those questions are narrower than “why did an AI model recommend a competitor?” The January data can show the sources Ahrefs records as cited and their frequency, but it does not by itself prove the causal reason for an answer, the quality of a recommendation, or a buyer’s likelihood of conversion.

That boundary matters in client reporting. A cited URL is useful evidence. It is not a complete explanation of model behavior.

Ahrefs Brand Radar API: cited domains in practical terms

A cited domain is a domain-level view of sources that appear in AI responses. If a publisher, review site, documentation site, or competitor domain repeatedly appears, the domain data can flag it as a source worth investigating.

Consider a fictional example for illustration only: an agency reviews AI answers around payroll-software questions and sees review-example.test repeatedly listed among cited domains. That is not an Ahrefs benchmark, a documented count, or proof that the review site determines every answer. It is a prompt to inspect the publisher’s relevant coverage and understand what information it makes available.

At domain level, teams can usually make three kinds of decisions.

1. Map the source landscape. Group cited domains into competitors, trade publications, review platforms, professional associations, forums, and other third-party publishers.
2. Identify recurring external sources. A frequently cited independent source may be more relevant to an outreach or review-management plan than another owned blog post.
3. Prioritize deeper page review. Domain data can tell a team where to look next; page data is needed to inspect the actual URL.

The January documentation supports cited-domain reporting with appearance frequency and estimated monthly search-demand context. It does not establish that a cited domain is authoritative, accurate, commercially influential, or universally used across every AI engine and prompt. Those are separate judgments that require human review.

A practical reporting note is to retain the reporting period, the tracked topic or prompt grouping used in the export, and the source of the data. Without that context, “most cited domain” can become a misleading headline because it hides the question set behind the result.

Cited pages are the URL-level follow-up

Cited pages provide a more specific object for content and competitive analysis: an individual URL rather than an entire site. That makes them particularly useful when a team needs to decide whether a competitor has a strong comparison page, whether an industry publication has a relevant guide, or whether an owned page needs factual updates.

The documented January additions allow teams to work with cited-page data and appearance frequency. The release material also refers to estimated monthly search demand connected to the associated mentions. We should not stretch that into claims that the endpoint delivers competitor topic clusters, buyer-intent labels, or source-quality scores; those fields are not established by the supplied January sources.

A responsible cited-page review process

When we use cited-page data in a workflow, we treat the API output as a research queue. A human still needs to read the page and determine whether it is relevant.

A simple working table can include:

• cited domain;
• cited URL;
• recorded appearance frequency;
• associated estimated monthly search demand, where available;
• source type, assigned manually, such as competitor, publisher, review site, or owned site;
• the page’s publication or update date, checked manually;
• an evidence note describing what the page actually contains; and
• a proposed action or a decision to take no action.

For example, a cited third-party comparison page may suggest that product details, pricing clarity, independent reviews, or category positioning need examination. It does not automatically mean that publishing a near-duplicate comparison page will change an AI answer.

This distinction protects teams from treating visibility data as a content brief generator. The API identifies observed citations; it does not prescribe the right editorial, PR, product-marketing, or technical SEO response.

What Share of Voice history can and cannot show

January’s overview-level Brand Radar API additions included AI Share of Voice distribution and historical data. This enables a team to programmatically incorporate a relative visibility measure into a recurring report rather than depending entirely on manual UI checks.

Share of Voice is most useful as a comparison within a stable reporting setup. If a report says one brand has a larger share than another for the same defined set of tracked data and time period, that is a useful directional observation. It is not the same as search traffic, market share, revenue share, or confirmed audience reach.

A careful monthly report could contain:

• the brand’s reported AI Share of Voice;
• the competitors included in the comparison;
• the reporting period;
• the change versus the prior comparable period;
• the highest-frequency cited domains and pages; and
• a short list of actions that the team actually completed.

Avoid attaching a causal story before the evidence exists. If Share of Voice changes after a page update, a new review, or a PR campaign, record the timing as a hypothesis to investigate. The January sources do not establish that a particular type of page update causes a particular Brand Radar movement.

For a broader way to frame reporting, our AI Visibility Index guide separates observed visibility from competitive context and business outcomes. That separation is especially useful when stakeholders expect a single percentage to answer every question.

Keep citations, mentions, and AI responses separate

AI visibility terminology often gets blurred, so reporting teams should define their terms before combining data sets. The January Ahrefs update is specifically relevant to cited pages and cited domains in AI-generated responses.

Here is a conservative taxonomy:

• AI-generated response: an answer produced by an AI system for a tracked query or request.
• Cited domain: a source domain recorded as cited in an AI response.
• Cited page: a specific source URL recorded as cited in an AI response.
• Brand mention: the brand name appears in answer text. This is a different signal from a citation.
• Share of Voice: a comparative visibility metric reported by the product; it should be presented with its reporting scope and date range.

A brand can be named without its own site being cited. Likewise, a domain can be cited as supporting information without the domain’s owner being the answer’s recommended product or service. These are reasons to avoid calling every citation a “win.”

The January materials supplied for this article do not document a citation-funnel system of states such as “found,” “found but not cited,” or “not found.” They also do not document API filters for those states. Teams should not backfill those labels into January 2026 reports as though they were part of this release.

Our comparison of AI citations versus AI visibility tracking explains why a source citation and a buyer-facing brand mention should remain separate measures.

What Page Types in Site Explorer added

The Site Explorer portion of the January update concerned backlink reports, not AI-response counts. Ahrefs added referring-page Page Types and page categories, along with columns and filters for those fields. The January announcement also states that this data is available through exports, the API, and Ahrefs MCP.

Ahrefs gives examples of Page Types such as articles, listicles, and tools. That is a format classification for referring pages in backlink reporting. It can help a team organize a backlink profile beyond a raw count of linking domains.

What Page Types can support

For a page with new backlinks, a team can use the fields to examine the formats of the referring pages. A cluster of editorial articles, listicles, or tool pages may lead to different follow-up questions about the kinds of assets attracting links.

For example, a company can filter a backlink report to inspect listicle-type referring pages, then review those pages manually for relevance and accuracy. The value is operational: it reduces the time needed to locate a particular page format in a large backlink report.

What Page Types do not establish

The supplied January source describes Page Types and page categories in backlink reports. It does not establish that a Page Type is a business-niche score, a measure of page authority, a quality rating, or an AI-citation classification.

It also does not say that an article, listicle, or tool page is inherently more likely to be cited by an AI system. Teams can compare backlink formats with cited-source observations in their own analysis, but they should label that comparison as an internal hypothesis rather than an Ahrefs product conclusion.

Brand Radar and Site Explorer answer different questions

The two January additions are complementary, but they should not be merged into one vague “AI visibility” score. Brand Radar’s cited-source data concerns pages and domains recorded in AI-generated responses. Site Explorer’s Page Types addition concerns the classification of referring pages in backlink reports.

The January source material does not establish Site Explorer columns for AI response counts per site or page. It also does not establish Grok availability in Brand Radar, Brand Radar API entity support, or platform-by-platform field behavior. Those claims should be sourced to their own dated announcements if a later article covers them.

This separation makes reports more credible. A backlink-format observation may be useful context for a content team, while cited-page data is direct evidence that a source appeared in the tracked AI-response dataset. Neither should be presented as proof of the other.

A cautious workflow for competitor-gap research

The release is most valuable when it feeds a repeatable decision process rather than a one-time list of cited sources. The following is our recommended workflow, not a claim about built-in Ahrefs automation or a universal benchmark.

1. Define the question set and reporting period

Start with buyer questions that reflect actual sales conversations, comparison needs, use cases, and problems. Record why each question belongs in the set. A small, carefully documented set can be more manageable than a large, unexplained keyword export, but no January Ahrefs source establishes an ideal prompt count.

2. Pull and label cited sources

Use Brand Radar cited-domain and cited-page data to identify the sources that occur in the recorded responses. Assign manual labels such as owned site, direct competitor, review platform, publisher, or association. Keep the original data separate from editorial interpretation.

3. Review the actual URLs

Read the cited pages. Check whether the page is current, whether it contains the claims relevant to the category, and whether it is a product page, editorial guide, comparison, tool, or another format. Page Type data in Site Explorer can provide backlink-report context, but it should not replace page review.

4. Write a testable gap hypothesis

A useful hypothesis is specific: “Competitor coverage includes implementation detail missing from our onboarding page,” or “Independent category pages cite information we have not published clearly.” A weak hypothesis is “we need more AI visibility.”

5. Make and log one concrete change

Update a factual page, create a needed comparison resource, correct a third-party listing, or decide that no action is warranted. Log the date, URL, owner, and intended question set. That record prevents later reports from inventing a causal explanation for a visibility change.

For teams that need to retain their own prompt-level tests alongside market-level source data, our AI search measurement framework outlines how we separate the questions we control from broader visibility reporting.

What the January update does not measure

The January 2026 release makes cited-source and Share of Voice reporting more accessible through the API. It does not replace business measurement or careful answer review.

It does not, based on the supplied sources, measure:

• conversions or revenue attributable to AI-generated answers;
• sentiment, factual accuracy, or recommendation quality in an answer;
• whether a cited source caused a brand mention;
• all possible brand-name variations and domains as a single API entity;
• citation-state filters for mentioned, uncited, or not-found results;
• Site Explorer counts of AI responses in which a website or page appears; or
• behavior differences across named AI platforms, including Grok.

This is not a weakness unique to Ahrefs. It is a reminder to match a metric to the decision it can support. Use cited pages and domains to investigate source patterns. Use Share of Voice history for comparative trend reporting. Use Page Types to organize backlink-report context. Use other measurement methods when the question is about prompt-level answer wording, business impact, or a controlled before-and-after test.

FAQ

What new Brand Radar API endpoints were released in January 2026?

Ahrefs’ January 2026 materials describe Brand Radar API additions for cited pages, cited domains, and overview data for AI Share of Voice distribution and history. The developer changelog dated January 2 also documents endpoint coverage for domains cited in AI responses. The cited-source data includes appearance frequency and associated estimated monthly search-demand context.

How can Brand Radar API endpoints filter mentions, citations, and uncited results?

The January 2026 source material used for this article does not document API filters for mention, citation-funnel, or uncited-result states. It supports cited-page and cited-domain reporting, not claims about filters such as “mentioned and cited,” “found but not cited,” or “not found.” Verify later dated documentation before adding those states to a report.

What are Page Types in Ahrefs Site Explorer, and what do they show?

Page Types are classifications for referring pages in Site Explorer backlink reports. Ahrefs’ January 2026 announcement gives articles, listicles, and tools as examples, and also introduced page categories, filters, columns, exports, API access, and Ahrefs MCP availability. They describe backlink-report page formats; they are not documented as AI-citation or authority scores.

How can Site Explorer report AI response and cited-page counts for a website?

The January 2026 materials distinguish these functions. Brand Radar provides the cited-page and cited-domain additions for AI-generated responses. Site Explorer’s January change was Page Types and categories in backlink reports. The supplied sources do not establish Site Explorer AI-response count columns for a website or page, so that capability should not be attributed to this release.

Can Brand Radar API combine brand-name variations and domains into one entity?

The January 2026 release material cited here does not document Brand Radar API entity support that combines name variations and domains into a single entity. Brand owners should therefore avoid assuming that normalization behavior from this release alone. If entity support is required, check Ahrefs’ dated developer documentation for the relevant endpoint and version before implementation.