Best AI Citation Tools for Tracking Sources Behind Brand Recommendations

Answer Capsule

Profound is the consensus leader for companies that need to track the sources behind AI brand recommendations, named by 5 of 6 platforms at an average listed position of 2.4 and a best position of 1. Peec AI is the strongest alternative for teams that want a clear used-versus-cited distinction and regional citation reporting, while OtterlyAI is the lowest-cost credible entry point at a published $29/month Lite tier. Ahrefs and Semrush suit buyers who want AI citation monitoring inside an existing SEO suite, and Slate and Scrunch are narrower, higher-variance options with conflicting public evidence. Six platforms were studied: OpenAI, Anthropic, DeepSeek, Grok, Perplexity, and Kimi. The principal limitation is that this index reflects one standardized prompt sent once to each platform, and platform answers vary by date, wording, location, model, and retrieved sources.

Research Snapshot

• Topic: AI citation and authority building — tools that reveal which sources sit behind AI brand recommendations.
• Target buyer: Companies seeking AI citation tools for tracking sources behind brand recommendations across AI search, generative-answer, and recommendation platforms.
• Platforms included: openai, anthropic, deepseek, grok, perplexity, kimi (6 platforms).
• Research date: 2026-09-17.
• Unique entities named: 27.
• Qualifying entities: 7.
• Eligibility rule: an entity had to be named by at least two platforms during ranking discovery.
• Ranking rule: platform mentions, then average listed rank, then best listed rank.
• Method: one standardized prompt sent once to each included platform.

Platform mentions count only ranking-discovery mentions. Several platforms later completed fit assessments for entities they did not name during discovery, so mention counts and fit-assessment counts are not interchangeable.

The Consensus Ranking

Questions This Section Answers

• What are the best AI citation tools for tracking sources behind brand recommendations in 2026?
• Which AI citation tracking tools were named by the most AI platforms in this study?

Which Option Is Best for Which Version of the Buyer Need?

Questions This Section Answers

• Which AI citation tool should a buyer choose if they need source-level citation mapping across the most AI answer engines?
• Which AI citation tracking tool is best for a buyer who needs a low-cost entry point before scaling prompt volume?
• Is Ahrefs or Semrush better for tracking AI citations if the buyer already runs an SEO suite?

1. Profound

Questions This Section Answers

• Is Profound worth it for tracking sources behind AI brand recommendations, and what are its main drawbacks?
• Does Profound distinguish between a brand citation and an active recommendation?
• What does Profound cost for enterprise AI citation tracking, and is billing annual only?

Profound is the consensus leader in this index and the most frequently named entity for source-level citation tracking behind AI brand recommendations. It was named by 5 of 6 platforms at an average listed position of 2.4, with a best position of 1. Its Answer Engine Insights product sends defined queries to answer engines, captures full responses and citation sources, and organizes prompts by topic, platform, region, and tags ^[1].

Why it ranked here. Profound received the top position from four of the five platforms that named it, and only Perplexity placed it lower, at 8. That spread reflects a real split in the evidence: platforms that could retrieve Profound's own documentation described a mature citation product, while platforms that could not verify the official site treated core capabilities as unconfirmed.

Best suited for. Enterprise brands and agencies monitoring recommendation prompts across ChatGPT, Perplexity, Gemini, Google AI Overviews, Microsoft Copilot, and other supported answer engines. It also suits teams that need source-level mapping, competitor benchmarking, citation-share trends, exports, and workflow links for content or outreach.

Main strengths for this use case. Profound reports cited URLs, citation frequency, citation share, citation coverage, source domains, and source categories including owned, competitor, earned media, PR wire, social, and institution, with ranked citation tables, exports, and citation-relationship visualization ^[2]. It tracks both domain-level and individual page-level citations ^[4], and its Sources dashboard surfaces the domains and URLs models cite when answering category questions ^[5]. Competitor citation tracking supports comparison by platform, topic, prompt, and region, including citation-share rankings ^[2]. Enterprise tracking is described as daily by default, with date-range comparisons, previous-period analysis, saved views, watched pages, and trend charts ^[6]. Profound also categorizes prompts by buying stage, distinguishing awareness-level queries from decision-stage queries ^[7], and it captures responses directly from the browser rather than API outputs ^[8].

Main limitations. The most consequential limitation is recommendation granularity. Independent reviews note that Profound aggregates mentions at surface level and does not distinguish between citations (a neutral mention in a list) and recommendations (an active top choice) ^[9]. Public materials also do not establish a validated recommendation-quality score or a causal model linking a source to a recommendation ^[1]. Coverage is sampled from prompts Profound runs, not a census of all real user recommendations ^[1]. Multiple independent reviewers flag a single-workspace architecture as a hard constraint for agencies and multi-brand teams ^[10]. Profound has no CMS integration, so citation data lives in the dashboard rather than connecting to publishing systems ^[12]. Independent reviews also report that Profound lacks first-class Reddit monitoring, despite Reddit driving a disproportionate share of AI citations ^[9]. Engine coverage is tier-dependent: the trial is ChatGPT-only, and the pricing page lists capability for ChatGPT, Perplexity, Google AI Mode, Google Gemini, Microsoft Copilot, DeepSeek, Anthropic Claude, Google AI Overviews, and Exa Search, but included engines depend on plan and configuration ^[13].

Pricing or cost summary. Public pricing is inconsistent across sources. Anthropic-reported evidence cites Starter at $99/month billed annually (ChatGPT only, 50 prompts, 1 seat, 100 Agent credits/month) and Growth at $399/month billed annually (3 engines including Perplexity and Google AI Overviews, 100 prompts, 3 seats, 400 Agent credits/month, unlimited exports), with annual-only billing and no month-to-month option ^[14]. Enterprise deployments are reported in independent reviews at $2,000–$5,000+ per month ^[17], while one Grok-reported review cites a $2,000–$8,000/month range ^[18]. Perplexity-reported summaries add a Lite tier at $499/month and enterprise pricing around $2,000–$5,000+ ^[19]. Profound Agents use a credit-based model, and the self-serve Agency Growth plan includes 400 credits per month per client workspace [e1:official:C2]. Pricing confidence across the evidence is low to moderate, and buyers should obtain a written quote.

Where platforms disagreed. OpenAI, Anthropic, Grok, and DeepSeek all placed Profound at 1. Perplexity placed it at 8 and rated the fit "good" rather than "strong," noting that citation and source-mapping depth for recommendations was not clearly verified in the available evidence ^[20]. Kimi rated the fit "uncertain," stating that Profound's public materials do not explicitly document extraction of exact cited source URLs per answer and that the platform appears oriented toward brand presence rather than forensic citation source mapping ^[21]. DeepSeek rated the fit "mixed" and reported that official-site retrieval failed during its research, leaving plan names and prices unverified ^[22]. Anthropic rated the fit "good" and flagged the recommendation-versus-citation gap as the key limitation ^[9].

The complete Profound fit review covers engine coverage, prompt volume, and contract questions in more detail.

2. Peec AI

Questions This Section Answers

• Is Peec AI better than Profound for tracking which sources AI models access versus which they cite?
• Which Peec AI plan should a buyer choose for multi-engine citation tracking, and what do extra engines cost?
• Does Peec AI attribute AI citations to traffic, leads, or pipeline?

Peec AI is the second-ranked entity and the strongest alternative for buyers who need a precise distinction between the sources a model accessed and the citations it displayed. It was named by 5 of 6 platforms at an average listed position of 2.6, with a best position of 2.

Why it ranked here. Peec AI was placed second by four of the five platforms that named it, with only Perplexity ranking it lower, at 5. Its distinguishing evidence is the used-versus-cited split, which directly addresses the question of which sources sit behind a recommendation rather than merely appearing in an answer.

Best suited for. Marketing, SEO, GEO, and content teams monitoring which domains and URLs influence brand recommendations, and companies comparing brand visibility, citation share, position, sentiment, and competitors across multiple AI engines. It also suits agencies or larger teams needing multi-project monitoring, reporting, API access, or custom model coverage.

Main strengths for this use case. Peec AI's sources and citations feature shows which URLs and domains an AI engine cited when forming its answer, distinguishing sources (all URLs accessed during response generation) from citations (URLs explicitly referenced in visible answer text) ^[23]. The Actions module identifies source gaps — domains and URLs that competitors are cited from but the user is not — and scores each gap by estimated impact ^[26]. The Actions engine clusters citation sources into owned, editorial, reference, and UGC categories and scores each opportunity 1–3 ^[27]. Peec AI runs prompts daily on every plan and compares week over week ^[28], and it is noted for region-based prompt execution and reporting, tracking AI citations per country or region ^[30]. It uses UI scraping rather than APIs for most tracked engines, capturing what a real user would see ^[32].

Main limitations. Peec AI has no built-in AI traffic estimation and no citation-to-lead attribution; if proving pipeline impact is the primary requirement, Peec alone will not provide it ^[34]. It identifies citation opportunities and gaps but does not generate the content needed to close them ^[36]. Starter and Pro plans restrict users to 3 AI models out of 6 available, with additional engines priced as add-ons ^[38]. Peec AI's terms state that it monitors and records source usage and citations but does not guarantee prompt suitability, accuracy, or effectiveness, and does not itself influence brand visibility ^[40]. Public materials do not establish that it performs a full technical citation-architecture audit of a buyer's website ^[41].

Pricing or cost summary. Anthropic-reported evidence cites Starter at €70/month annual or about €82 monthly (50 prompts, 1 project, 3 AI models), Pro at €180/month annual or about €212 monthly (150 prompts, 2 projects), and Advanced at €360/month annual or about €420 monthly (350 prompts, 5 projects), with USD pricing around $95–$420 for corresponding tiers ^[42]. Additional AI engines cost €20–30 per engine per month ^[44]. All plans include unlimited seats, Peec MCP, API, all base AI channels, Looker Studio, CSV exports, and daily tracking ^[45]. A 7-day free trial with no credit card required was reported ^[46]. Capterra listings show different figures — $95 Starter, $245 Pro, $495 Advanced — which suggests regional pricing variation or outdated directory data ^[43]. The reviewed official pricing page exposed Starter, Pro, Advanced, and Enterprise tiers without reliable displayed monetary amounts ^[47].

Where platforms disagreed. OpenAI, Anthropic, Grok, and Perplexity rated Peec AI "good" or "strong," while DeepSeek rated it "mixed" and Kimi rated it "uncertain." DeepSeek reported that official-site content retrieval did not succeed, so capability and pricing claims rest mainly on company-level and third-party listing descriptions ^[48]. Kimi stated that no independent source documentation of Peec AI's actual citation-tracking features, pricing, or engine coverage was found in its reviewed materials ^[49]. Perplexity noted that public materials do not clearly prove full source-mapping depth such as answer-by-answer provenance graphs ^[50]. The number of available AI engines is also inconsistent: Peec's pricing page lists up to 11 LLM models on Enterprise while the AI-instructions page lists 13 ^[51].

The complete Peec AI fit review covers plan entitlements, engine add-ons, and attribution gaps.

3. OtterlyAI

Questions This Section Answers

• Is OtterlyAI worth it for tracking AI citations behind brand recommendations, and what are its main drawbacks?
• Which OtterlyAI plan has the lowest published cost, and what do extra engines and prompts add?
• Does OtterlyAI label a response as a recommendation, or only as a mention or citation?

OtterlyAI is the lowest-cost credible entry point in this index and the third-ranked entity overall. It was named by 5 of 6 platforms at an average listed position of 3.6, with a best position of 3.

Why it ranked here. OtterlyAI was placed third by three platforms and fourth or fifth by the others, producing a tighter spread than the two entities above it. Its ranking reflects broad agreement that it monitors citations and sources at a low entry price, combined with shared concern about engine add-on costs and refresh cadence.

Best suited for. Marketing and SEO teams tracking brand recommendations and citations across ChatGPT, Google AI Overviews, Perplexity, and Microsoft Copilot, with Claude, Gemini, and Google AI Mode available as paid add-ons. It also suits teams needing source-level citation reports, competitor comparisons, historical change monitoring, and exportable dashboards.

Main strengths for this use case. OtterlyAI parses URLs and domain citations from AI responses and tracks citation frequency, position changes, and domain rankings over time across monitored engines ^[52]. Its Citations report supports date, tag, engine, country, domain, and URL filtering, competitor gap analysis, and daily updates ^[53]. The platform captures citation links and provides Link Citations Analysis, citation details, URL and domain filtering, winners and losers, and source-level monitoring ^[54]. It queries public AI interfaces rather than relying only on APIs, captures answer text and citation links, and warns that results can differ from a logged-in or personalized session ^[55]. GEO Audit evaluates 25+ on-page factors affecting citation likelihood and provides page-level recommendations ^[56]. Standard and Premium tiers add API, MCP, Looker Studio connector, CSV export, and PDF reporting ^[57].

Main limitations. Public documentation does not prove causal attribution for why a brand was recommended ^[54]. An independent review cautions that being mentioned, being recommended, and being used as a source should not automatically be treated as equivalent ^[58]. Multiple independent reviews report a weekly refresh cycle with up to a 7-day monitoring lag, which is problematic for rapidly evolving topics ^[59]. The included engine set is limited, and important engines require add-ons ^[60]. Cancellation may result in deletion of tracked engines and historical data, creating a retention risk for longitudinal programs ^[61]. No retroactive analysis is available, so new accounts cannot establish historical baselines ^[62]. Sentiment analysis is reported as less developed than core monitoring features, with accuracy varying by platform ^[63].

Pricing or cost summary. The public pricing page lists Lite at $29/month (15 prompts, one workspace), Standard at $189/month (100 prompts, unlimited workspaces, API, MCP, and agent analytics allowances), and Premium at $489/month (400 prompts, higher allowances), with annual billing advertised at a 15% discount [e3:official:C2]^[60]. Additional 100 prompts cost $99/month on Standard and Premium [e3:official:C2]. Monthly engine add-ons are listed at $9/$59/$149 for Google AI Mode and Gemini and $29/$109/$439 for Claude across Lite/Standard/Premium respectively [e3:official:C2]. Enterprise starts from $1,000/month with SSO and a dedicated CSM ^[65]. Applicable taxes are excluded from displayed US-dollar prices ^[60].

Where platforms disagreed. OpenAI, Anthropic, Grok, and Perplexity rated the fit "good," while DeepSeek rated it "mixed" and Kimi rated it "uncertain." DeepSeek noted that public pricing conflicts across sources and that whether source mapping is answer-level or tied to specific recommendations remains unclear ^[66]. Kimi stated that no verified public evidence confirms OtterlyAI delivers granular URL-level citation tracking, source mapping, citation architecture analysis, and historical citation decay monitoring ^[67]. Anthropic reported that sentiment analysis accessibility is inconsistent across independent reviews, with some reporting it is not accessible in the main dashboard ^[64]. The public pricing page states multi-country support of 50+ while a help page states 65+ ^[68].

The complete OtterlyAI fit review covers prompt quotas, engine add-ons, and data-retention terms.

4. Ahrefs

Questions This Section Answers

• Is Ahrefs Brand Radar worth it for tracking which pages and domains are cited behind AI brand recommendations?
• How much does Ahrefs Brand Radar cost per engine versus all platforms, and is a base Ahrefs subscription required?
• Does Ahrefs expose the exact source URLs cited by AI engines, or only brand mentions?

Ahrefs is the fourth-ranked entity and the strongest option for buyers who want AI citation monitoring inside an established SEO and backlink workflow. It was named by 5 of 6 platforms at an average listed position of 4.6, with a best position of 4.

Why it ranked here. Ahrefs was placed fourth by three platforms and fifth or sixth by the others. Its ranking reflects broad recognition of Brand Radar's citation and cited-page reporting combined with consistent concern about monthly refresh cadence, add-on pricing, and dependency on the wider Ahrefs suite.

Best suited for. Marketing, SEO, and AEO teams tracking which pages and domains are cited behind brand or product recommendations, and companies comparing brand mentions, citations, AI Share of Voice, and competitors across multiple AI answer platforms. It also suits teams combining broad market discovery with custom monitoring of high-intent recommendation prompts.

Main strengths for this use case. Brand Radar separates brand mentions from citations and defines citations as linked pages or domains used as sources in AI-generated responses, and it distinguishes pages that were found but not cited ^[69]. It reports AI Share of Voice, mentions, citations, and impressions relative to competitors ^[71]. Custom Prompts lets buyers define exact questions about products, brands, competitors, and categories and monitor responses on Google AI Overviews, Google AI Mode, ChatGPT, Perplexity, Gemini, Copilot, and Claude where supported ^[72]. Ahrefs states that the AI Visibility Index provides historical responses back to 2025 ^[73]. The API exposes citation fields and page-status classifications, and custom-prompt-only requests are documented as free for that endpoint ^[70]. Brand Radar also connects AI citation data to Domain Rating, backlinks, branded web mentions, and branded search volume ^[75].

Main limitations. The broad index is modeled from Ahrefs search-backed prompts and may not represent every real buyer question or personalized response ^[69]. Citation visibility is not causal attribution; public documentation does not show why a particular source caused a recommendation ^[69]. Custom Prompt checks are usage-metered by prompt, model, and location, which can make multi-market, multi-platform monitoring expensive ^[72]. Grok availability is described inconsistently across Ahrefs pages published at different times, and one help page notes that following recent policy changes by Grok, Ahrefs was temporarily unable to gather new data ^[76]. AI Visibility Index coverage is limited for brands with low search volume ^[77]. Estimated Impressions are modeled on Google search volume, and Ahrefs documentation explicitly states it does not claim a validated link between Google search volume and how often a query is asked inside an AI tool ^[78].

Pricing or cost summary. Public Ahrefs pages describe Brand Radar AI Visibility Index access starting at $199/month for a single platform/index and $699/month for all platforms, with Custom Prompts separately described as starting at $50/month ^[79]. Custom Prompt packages are listed at Basic $50/month for 2,500 checks, Growth $100/month for 7,000 checks, and Scale $250/month for 25,000 checks, with documented overage rates of $0.020, $0.015, and $0.010 per check respectively ^[81]. A check is calculated as one prompt execution multiplied by one LLM and one location ^[81]. Ahrefs states Brand Radar can generally be purchased monthly, with one pricing article saying only Enterprise users can buy it annually ^[82]. One independent review reports a conflicting standalone price of $398/month for select platforms ^[83].

Where platforms disagreed. OpenAI, Anthropic, and Grok rated the fit "good," while Perplexity and Kimi rated it "mixed" and DeepSeek rated it "weak." DeepSeek stated that Brand Radar tracks brand mentions and visibility but does not expose the underlying source URLs that AI engines used, calling this a fundamental gap for citation tracking ^[84]. This directly conflicts with OpenAI- and Anthropic-reported evidence that Brand Radar distinguishes citations from found-but-not-cited pages and exposes cited pages and cited domains ^[69]. Kimi reported that Brand Radar appears to prioritize brand mention detection over granular URL-level citation tracking ^[86]. Perplexity noted that public materials do not verify a complete source-lineage or citation-architecture analysis workflow for every answer ^[87].

The complete Ahrefs fit review covers index coverage, custom prompt metering, and platform availability.

5. Semrush

Questions This Section Answers

• Is Semrush AI Visibility Toolkit worth it for tracking sources behind AI brand recommendations, and what does it miss?
• How much does Semrush AI Visibility Toolkit cost per domain, and what happens to cost across multiple brands?
• Does Semrush track Claude, Perplexity, and Grok citations in its base plan?

Semrush is the fifth-ranked entity and the best fit for buyers who want AI citation monitoring integrated with conventional SEO, site auditing, reporting, and content workflows. It was named by 4 of 6 platforms at an average listed position of 4.75, with a best position of 3.

Why it ranked here. Semrush received a best position of 3 from DeepSeek but was placed fifth, fourth, and seventh by the other naming platforms. Its ranking reflects strong integration value for existing Semrush customers alongside consistent evidence of engine coverage gaps and per-domain cost escalation.

Best suited for. Marketing and SEO teams benchmarking brand recommendations and citations across ChatGPT, Google AI, Gemini, and Perplexity, and companies seeking competitor and source-gap analysis rather than only raw citation exports. It also suits organizations that want AI citation monitoring integrated with conventional SEO, site auditing, reporting, and content workflows.

Main strengths for this use case. The toolkit reports AI visibility, brand mentions, citations, cited pages, cited sources, and source opportunities, and it distinguishes mentions (unlinked brand references) from citations (mentions with linked sources) ^[88]. It maps external websites and company pages cited in AI answers and identifies sources cited for competitors but not the buyer ^[90]. Competitor Research supports side-by-side comparison of mentions, citations, topic coverage, average citation position, and missing sources for up to four competitors ^[91]. Visibility Overview provides AI-visibility trends, with daily rolling updates for prompt-database data, weekly updates for Brand Performance, and daily updates for custom Prompt Tracking ^[92]. Documented coverage includes ChatGPT, Google AI Overviews, Google AI Mode, Gemini, and Perplexity, with regional analysis across 117 regional databases ^[93]. The toolkit supports PDF and CSV exports, scheduled reports, shareable dashboards, and My Reports integration, with CSV exports limited to 1,000 rows per export and 10 exports per user per day ^[94].

Main limitations. Base pricing includes only 25 custom prompts with daily AI rankings and one Brand Performance domain ^[95]. Claude is excluded from the base plan and only available in custom Enterprise AIO, and Perplexity coverage is inconsistent across reports ^[96]. Metrics are directional estimates and may not reproduce every user-specific or personalized AI answer ^[93]. One 2026 independent test found sentiment classification carries roughly 6% disagreement with manual coding, and Google AI Mode metrics carried beta labels as of May 2026 ^[98]. Coverage is US-English-centric with around six regional databases drawing on US-English data ^[97]. The toolkit is monitoring-only and does not write, optimize, or refresh content ^[97]. Independent sources also report bugs in competitor configuration ^[99].

Pricing or cost summary. Semrush lists AI Visibility Toolkit Base at $99 per month per domain when billed annually, covering one domain, 25 tracked prompts, and one user seat ^[95]. Semrush One bundles are listed at Starter $199/month, Pro+ $299/month, and Advanced $549/month, with annual billing applying a 17% discount ^[101][e5:official:C2]. Additional domains cost $99/month each, and a 10-domain agency portfolio runs roughly $1,090/month for the toolkit alone ^[102]. Enterprise AIO pricing is custom and not publicly specified ^[95]. There is no free trial for the standalone toolkit, though Semrush One includes a 7-day trial ^[101].

Where platforms disagreed. OpenAI, Anthropic, and DeepSeek rated the fit "good," while Grok and Perplexity rated it "mixed" and Kimi rated it "weak." Kimi stated that Semrush does not log the specific source URLs that AI engines cite alongside brand recommendations and does not classify cited sources by type ^[104]. This conflicts with Anthropic- and OpenAI-reported evidence that the toolkit tracks citations at page level and captures which specific URLs are cited ^[105]. Engine coverage statements also conflict: independent reviews from mid-2026 report Gemini as rolling out and Perplexity or Claude limited or absent, while Semrush's own documentation lists ChatGPT, AI Overviews, AI Mode, Gemini, and Perplexity ^[97]. Grok noted that methodology transparency is limited, with reviewers describing black-box scoring ^[106].

The complete Semrush fit review covers per-domain stacking, engine tiers, and export limits.

6. Slate

Questions This Section Answers

• Is Slate worth it for tracking AI citations behind brand recommendations, and what are its main drawbacks?
• What does Slate cost per month, and does the published pricing match the plan names cited in third-party roundups?
• Does Slate verify whether a cited source actually supports the AI recommendation it appears alongside?

Slate is the sixth-ranked entity and the highest-variance option in this index. It was named by only 2 of 6 platforms, but one of those platforms placed it first, producing a best position of 1 and an average listed position of 4.0.

Why it ranked here. Slate's ranking is driven by a single top placement from Perplexity combined with a seventh-place placement from Anthropic. That split is visible in the evidence: platforms that retrieved Slate's own documentation described a functioning citation-analysis product, while platforms that could not verify the official site treated the product as unproven.

Best suited for. Marketing, SEO, content, and agency teams monitoring brand recommendations across multiple AI answer engines, and companies that want citation-source discovery combined with prompt history, competitor visibility, content optimization, and outreach or workflow execution.

Main strengths for this use case. Slate's Citation Analysis engine tracks which URLs and domains AI platforms cite when generating brand mentions, categorizing citations into My Pages, Competitor Pages, Social Pages, and Other Pages with domain-level breakdowns showing frequency, influence, and platform attribution ^[107]. It reports brand mentions, average position in recommendations, visibility scores, sentiment, competitor share of voice, and prompt-level results across AI answer platforms ^[108]. It captures the full, unedited transcript of every AI interaction with a searchable archive of every prompt run ^[109]. Citation Trends charts show how owned citations, mentioned citations, and total citations change over time ^[107]. Slate publicly lists tracking across ChatGPT, Perplexity, Google AI Overviews, Google AI Mode, Claude, and Gemini, with Claude and Gemini as add-ons ^[108]. It also connects monitoring to workflows that can research, draft, refresh, and publish citation-oriented content ^[108].

Main limitations. Slate requires pre-configuration of tracked prompts and does not automatically capture all brand mentions or recommendations in real time ^[109]. It does not provide technical analysis of citation architecture, model-level source attribution, citation verification, or audit trails for citation accuracy ^[109]. It does not independently verify whether cited sources actually contain the information AI systems claim ^[109]. Sentiment analysis is provided at response level, not at individual citation level ^[109]. Public evidence is primarily vendor-reported, and independent testing of source coverage and recommendation accuracy was not found ^[108]. The public materials do not clearly establish API availability, raw citation-data access, retention duration, per-prompt limits, source deduplication rules, or model-version locking ^[108].

Pricing or cost summary. The current public pricing page lists Growth at $499/month (15,000 workflow credits, 10,000 AI answers tracked per month, 1,000 monitored pages), Scale at $999/month (50,000 workflow credits, 20,000 AI answers, 5,000 monitored pages), and Agency at $2,000/month (100,000 workflow credits, 50,000 AI answers, 10,000 monitored pages), with Enterprise custom and a 14-day trial ^[110]. Additional workflow credits are listed at $125 per 5,000 credits, and Claude and Gemini are listed as add-ons at $100 per 1,000 answers ^[110]. Annual billing is shown as 20% off ^[110]. Third-party sources cite a $199/month starting price and a $5,000–$20,000 per year enterprise range, neither of which was confirmed on the current official pricing page ^[111].

Where platforms disagreed. Grok rated the fit "strong," OpenAI and Anthropic rated it "good," DeepSeek and Perplexity rated it "mixed," and Kimi rated it "uncertain." Kimi stated that no verified evidence confirms Slate operates an AI citation tracking tool, that the entity name conflicts with other "Slate"-branded products, and that the provided official website URL did not appear in any supplied search results about AI citation tracking tools ^[113]. DeepSeek reported that no independent documentation of citation-tracking, source-mapping, recommendation-level data, or historical AI-answer monitoring was found ^[114]. Perplexity noted that public pricing is inconsistent, with Slate's own pages and third-party listings showing $199/month while other Slate pages describe pricing on request or list Solo at $199/month and Pro at $499/month ^[115]. The supplied Professional and Team plan labels could not be verified on the current pricing page ^[110].

The complete Slate fit review covers plan identity, engine add-ons, and raw-data access.

7. Scrunch

Questions This Section Answers

• Is Scrunch worth it for enterprise tracking of sources behind AI brand recommendations, and what are its main drawbacks?
• What does Scrunch Enterprise cost, and is pricing published anywhere?
• Does Scrunch verify citation accuracy, or only track citation consistency and presence?

Scrunch is the seventh-ranked entity and the strongest enterprise-governance option in this index. It was named by only 2 of 6 platforms, with a best position of 2 and an average listed position of 4.0.

Why it ranked here. Scrunch received a second-place placement from Perplexity and a sixth-place placement from Anthropic. Its ranking reflects strong enterprise feature evidence from platforms that retrieved its documentation, offset by limited naming across the platform set and unresolved questions about pricing and post-acquisition product direction.

Best suited for. Enterprise brands monitoring how ChatGPT, Claude, Gemini, Perplexity, Google AI surfaces, Meta AI, Microsoft Copilot, and Grok cite their brand, competitors, and third parties. It also suits teams needing prompt-level answers, cited URLs, source-owner classification, competitor benchmarking, and trend reporting, and organizations requiring custom prompt volumes, multiple workspaces, API access, SSO, integrations, and dedicated support.

Main strengths for this use case. Scrunch tracks citations at both domain level and URL level, recording the full list of URLs cited, which pages contributed to AI model answers, whether the brand appears in those pages, and topical categorization ^[117]. Its Citations tab provides Group by Domain and Group by URL views, with prompt-level performance, platform filters, cited URLs, and brand or competitor presence ^[118]. It calculates an Influence Score as Citation Consistency multiplied by Number of Unique Prompts, measuring how broadly and consistently a source shapes AI responses ^[119]. Citations can be filtered by owner — Owned, Competitor, Third Party — allowing segmentation by pages the buyer controls, competitor domains, and external sites worth pitching ^[117]. Date-based filtering defaults to the last 12 weeks with visual trend charts ^[117]. The Enterprise pricing page lists nine supported AI platforms: ChatGPT, Claude, Perplexity, Gemini, Meta AI, Google AI Mode, Google AI Overviews, Microsoft Copilot, and Grok ^[120]. Scrunch holds SOC 2 Type II certification and offers SSO, API access, and multi-brand workspaces on Enterprise ^[121].

Main limitations. Public evidence is primarily vendor-owned, and independent validation of citation completeness, accuracy, and recommendation-level outcomes is limited ^[122]. Scrunch warns that JavaScript-only pages, bot blocking, and temporary retrieval errors can prevent access to full cited-page content ^[123]. Public materials do not verify causal source attribution, citation correctness, or a formal citation graph or architecture audit ^[122]. Enterprise price and commercial terms are undisclosed ^[120]. Core tier is limited to 125 unique prompts, one brand workspace, and one country, which constrains agencies and multi-market organizations ^[124]. Scrunch does not independently verify citation accuracy, source credibility, or detect AI hallucinations; quality assessment relies on user manual review ^[125]. Real-time streaming of citation changes is not documented as a self-serve capability ^[125].

Pricing or cost summary. Enterprise pricing is custom and not publicly stated ^[120]. Scrunch publicly lists Core at $250 per month for 125 unique prompts, one brand workspace, five user licenses, five site audits per month, and four listed AI platforms, with a 7-day Core trial ^[126]. Annual billing includes a 17% discount ^[127]. Extra seats are reported at roughly $25/month per additional seat ^[121]. Potential ongoing cost drivers include prompt volume, number of workspaces, user licenses, countries, languages, competitors, model coverage, integrations, and support scope, with exact scaling rules unclear ^[120]. The Cloudflare Enterprise Logpush integration is described as having no additional Cloudflare or Scrunch cost for that integration path ^[128].

Where platforms disagreed. OpenAI, Anthropic, and Grok rated the fit "strong," while DeepSeek and Perplexity rated it "good" and Kimi rated it "uncertain." Kimi stated that Scrunch appears to be a legacy influencer marketing platform absorbed into Sitecore's enterprise marketing stack, with no verifiable current capability in this use case, and that no 2024–2026 independent reviews confirm Scrunch as an active competitor in AI citation tracking ^[129]. This conflicts with OpenAI- and Anthropic-reported evidence describing a functioning Citations tab, Influence Score, and nine-platform coverage ^[122]. DeepSeek reported that citation-level source mapping depth and methodology are not publicly documented ^[130]. Perplexity noted that public documentation does not clearly disclose historical monitoring depth or retention ^[131]. The Sitecore acquisition is reported at approximately $225 million, though the financial terms are not confirmed by the companies ^[132].

The complete Scrunch fit review covers Enterprise terms, platform coverage, and post-acquisition roadmap questions.

What the Cross-Platform Study Reveals About This Market

Questions This Section Answers

• What do AI platforms consistently look for in an AI citation tracking tool for brand recommendations?
• Which capabilities separate the top-ranked AI citation tools from the rest of the field?

Three capabilities recur across nearly every platform's reasoning. The first is source-level granularity: the ability to report the specific URLs and domains cited in an AI answer, not just whether a brand was mentioned. Profound, Peec AI, OtterlyAI, Ahrefs, Semrush, Slate, and Scrunch all received credit for some version of this, and the platforms that withheld credit did so on the grounds that the capability was unverified rather than absent.

The second is the distinction between a mention, a citation, and a recommendation. Peec AI received the clearest credit for separating sources accessed from citations displayed ^[133]. Ahrefs received credit for separating citations from found-but-not-cited pages ^[134]. Profound was penalized by Anthropic for not distinguishing neutral citations from active top-choice recommendations ^[135]. OtterlyAI was flagged for the same gap by an independent review ^[136]. This distinction is the single most contested capability in the category.

The third is historical monitoring. Profound collects citation data daily with complete response snapshots ^[137]. Peec AI runs prompts daily and compares week over week ^[138]. OtterlyAI supports date-range filtering on its Citations report ^[139]. Ahrefs stores historical AI responses back to 2025 ^[140]. Semrush documents daily rolling updates for prompt-database data ^[141]. Scrunch provides date-based filtering with a 12-week default lookback ^[142].

A fourth pattern is that no platform in this study treated any vendor's citation data as ground truth. Independent academic research cited in the Profound bundle documents frequent hallucination, inaccurate citation, and variation in answer-engine behavior, supporting caution when interpreting citation-monitoring metrics as ground truth ^[143]. Peec AI's own terms state that it does not guarantee prompt suitability, accuracy, or effectiveness ^[144]. Semrush states that AI visibility metrics are directional because AI responses are changing and personalized ^[145].

Where the AI Platforms Agreed

Questions This Section Answers

• Which AI citation tracking tools did every platform agree belong on a shortlist for brand recommendation monitoring?

All six platforms that completed fit assessments agreed that Profound, Peec AI, OtterlyAI, Ahrefs, Semrush, Slate, and Scrunch are legitimate candidates for this use case, even when they disagreed on fit strength. No platform argued that any of the seven should be excluded outright.

There was broad agreement that source-level citation tracking is the defining requirement. OpenAI, Anthropic, Grok, and Perplexity all credited Profound with citation and source-mapping capability ^[146]. OpenAI, Anthropic, Grok, and Perplexity all credited Peec AI with source and citation tracking ^[149]. OpenAI, Anthropic, Grok, and Perplexity all credited OtterlyAI with citation and URL tracking ^[152].

There was also agreement that competitor comparison is a core feature rather than a nice-to-have. Profound, Peec AI, OtterlyAI, Ahrefs, Semrush, Slate, and Scrunch all received credit for competitor benchmarking or share-of-voice comparison from at least three platforms.

Finally, platforms agreed that pricing transparency is a widespread problem in this category. Profound, Peec AI, Ahrefs, Semrush, Slate, and Scrunch all had pricing conflicts or gaps flagged. Only OtterlyAI had pricing described as high-confidence across multiple platforms ^[155].

Where the AI Platforms Disagreed

Questions This Section Answers

• Why did some AI platforms rank Profound, Ahrefs, Semrush, Slate, and Scrunch lower than others?
• Which AI citation tool rankings were most contested across platforms in this study?

The deepest disagreement concerns whether specific vendors actually expose cited source URLs. DeepSeek rated Ahrefs "weak" and stated that Brand Radar does not expose the underlying source URLs that AI engines used ^[158]. OpenAI and Anthropic reported the opposite, describing cited pages, cited domains, and a found-but-not-cited classification ^[159]. Kimi made a similar claim about Semrush, stating that it does not log the specific source URLs cited alongside brand recommendations ^[161], while Anthropic and OpenAI reported page-level citation tracking ^[162].

A second disagreement concerns whether vendors distinguish recommendations from citations. Anthropic stated that Profound does not parse whether mentions are neutral citations or active top-choice recommendations ^[164]. OpenAI described the same gap as unclear rather than absent ^[165]. Grok rated Profound "strong" without raising the issue ^[166].

A third disagreement concerns entity verification. Kimi rated Slate "uncertain" and stated that no verified evidence confirms the company operates in this category ^[167], while Grok rated Slate "strong" and described detailed citation-tracking features ^[168]. Kimi made a parallel claim about Scrunch, describing it as a legacy influencer marketing platform ^[169], while OpenAI, Anthropic, and Grok described a functioning enterprise citation product ^[170].

A fourth disagreement concerns engine coverage. Semrush's own documentation lists ChatGPT, AI Overviews, AI Mode, Gemini, and Perplexity, while independent reviews from mid-2026 report Gemini as rolling out and Perplexity or Claude limited or absent ^[173]. Ahrefs pages describe Grok availability inconsistently, and one help page notes a temporary pause on new Grok data ^[174].

How Buyers Should Choose

Questions This Section Answers

• What should a buyer check before choosing an AI citation tool for tracking sources behind brand recommendations?
• Which AI citation tracking tool should a buyer pick if they need verified source URLs rather than aggregate visibility scores?

Start with the output format, not the feature list. Ask each vendor to show a live example of a single tracked prompt with the full answer text, every cited URL, the citation position, the timestamp, the engine, the model, and the region. If a vendor cannot produce that artifact, the citation-tracking claim is unverified regardless of how the dashboard looks.

Second, decide whether you need mentions, citations, or recommendations. These are three different measurements. Peec AI and Ahrefs received the clearest credit for separating sources from citations ^[175]. No vendor in this study received unambiguous credit for identifying an active top-choice recommendation as distinct from a neutral citation.

Third, calculate total cost including add-ons. OtterlyAI's $29/month Lite tier rises quickly once Claude, Gemini, and Google AI Mode are added at $9–$439/month depending on tier [e3:official:C2]. Semrush's $99/month per-domain price becomes roughly $1,090/month for a 10-domain portfolio ^[177]. Ahrefs' $199/month single-platform price becomes $699/month for all platforms, plus Custom Prompt packages ^[178].

Fourth, verify engine coverage against your actual buyer queries. Profound's trial is ChatGPT-only ^[179]. Semrush's base plan excludes Claude ^[180]. OtterlyAI's base plans include four engines with three more as add-ons ^[181]. Ahrefs documents Grok collection limitations ^[182].

Fifth, confirm data retention and export terms before signing. OtterlyAI's cancellation help page states that tracked engines and historical data are deleted after account cancellation ^[183]. Semrush limits CSV exports to 1,000 rows per export and 10 exports per user per day ^[184]. Slate's public materials do not specify retention duration ^[185].

Buyers who want the full set of category-level comparisons across vendors can browse the ai citation authority building directory.

Methodology

This index was produced from a single standardized prompt sent once to each of six included platforms on 2026-09-17: OpenAI, Anthropic, DeepSeek, Grok, Perplexity, and Kimi. The prompt asked which AI citation or visibility tools the platform would recommend to a company that wants to understand not only whether AI systems recommend its competitors but which sources appear in the same answers and may contribute to those recommendations, and why.

Platforms named 27 unique entities. An entity qualified for the final ranking if it was named by at least two platforms during ranking discovery. Seven entities qualified.

The final ranking order is determined by platform mentions, then average listed rank, then best listed rank. The ranking table is the sole authority for rank, platform mentions, platform share, average listed position, and best position. Platform mentions count only ranking-discovery mentions and do not include platforms that later completed a fit assessment without naming the entity during discovery.

Entity evidence bundles supplied the buyer-fit, feature, pricing, limitation, and disagreement details used in each ranked section. Citations are platform-reported evidence, not independently verified facts. Company-owned sources are distinguished from independent sources where the evidence bundles provided that information. Several entities had materially more company-owned citations than independent citations, and company claims are not described as independently verified anywhere in this report.

Methodology Limitations

This study used one standardized prompt sent once to each included platform. AI answers can vary by date, wording, location, account state, model, interface, browsing configuration, and the sources retrieved. A different prompt, a different day, or a different account state could produce a different ranking.

Platform-reported research dates differ from the authoritative run date of 2026-09-17. DeepSeek reported 2026-02-14 for Profound, 2026-01-15 for Peec AI, 2026-06-12 for OtterlyAI, 2026-09-17 for Ahrefs, 2026-01-15 for Semrush, 2026-01-15 for Slate, and 2026-06-11 for Scrunch. Anthropic reported 2025-01-16 for Ahrefs, 2026-01-15 for Slate, and 2026-01-15 for Scrunch. These platform-reported dates are provenance metadata and do not independently prove freshness.

Official-site retrieval failed for one or more mentions of Profound, Peec AI, Ahrefs, and Scrunch. No failed fetch was used as a verified domain key. Where official-site retrieval failed, identity was matched by exact-name fallback, and the matching reported domain remains unverified.

The supplied URLs were collected from platform responses and were not independently validated by the writer stage. Citations are platform-reported evidence, not independently verified facts. No-search model claims require explicit verification before being described as current facts.

Platform recommendations are market intelligence, not independent customer reviews or proof of quality. This index does not reward or penalize any entity based on affiliate status.

Final Verdict

Profound is the consensus leader for tracking the sources behind AI brand recommendations, named by 5 of 6 platforms at an average listed position of 2.4. It is the strongest single choice for enterprise teams that need source-level citation mapping, competitor citation-share comparison, and daily historical monitoring across major answer engines, provided they can accept annual billing, a single-workspace architecture, and the absence of a validated recommendation-versus-citation distinction.

Peec AI is the strongest alternative for buyers who need a clear used-versus-cited distinction and regional citation reporting, and it is the better choice when source-gap analysis matters more than content execution. OtterlyAI is the best entry point for buyers who want published, transparent pricing and can accept a weekly refresh cadence and engine add-on costs. Ahrefs and Semrush are the right choices when AI citation monitoring must live inside an existing SEO workflow, with Ahrefs stronger on citation-versus-found classification and Semrush stronger on competitor and source-gap reporting. Slate and Scrunch are higher-variance options: both received strong placements from at least one platform and unverified or uncertain assessments from others, and both require direct vendor validation of pricing, plan identity, and citation-level output before purchase.