Best AI visibility MCP servers: check brand mentions from Claude and ChatGPT

Several MCP servers let Claude, ChatGPT, Codex or Cursor check how your brand shows up in AI answers. Cite42 runs new checks on any prompt across ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews and charges per call, with no subscription. Ahrefs (through Brand Radar) and SE Ranking include AI visibility tools in their SEO MCP servers. OtterlyAI, Peec AI, LLM Pulse and Profound connect your assistant to the monitoring data in their dashboards.

One question sorts them: should the assistant run a new check right now, or read data a platform has already collected? This guide compares what each server's tools do, which AI engines they cover, how you sign in and what it costs as of October 2026.

Cite42 publishes this guide and runs one of the servers in it. Everything about the other servers comes from their official MCP documentation and pricing pages, checked on October 5 and 6, 2026, rather than from paid trials.

The best AI visibility MCP servers for different jobs

• Cite42: fresh brand, citation and sentiment checks on your own prompts from inside the assistant, paid per call with no subscription.
• Ahrefs MCP: querying Brand Radar's index of more than 400 million prompts for any brand, if you have an Ahrefs plan.
• SE Ranking: AI search visibility next to keyword, backlink and audit tools, on pay-as-you-go API credits.
• OtterlyAI: reading and extending a daily monitoring program, including adding prompts, from Claude.
• Peec AI: reading Peec reports and managing prompts, topics and brands in batches.
• LLM Pulse, Profound and Scrunch: assistant access to their monitoring platforms, from LLM Pulse's €49 Starter plan to Profound's Enterprise contract.

Comparison at a glance

We built the shortlist from the servers AI assistants already recommend. On October 6, 2026 we asked ChatGPT, Perplexity and Gemini which MCP servers let Claude or ChatGPT check how a brand shows up in AI answers. They named servers from Peec AI, Profound, OtterlyAI, LLM Pulse and Cite42, plus about a dozen community projects on GitHub and npm. We kept the hosted servers whose official docs describe AI visibility tools, and added Ahrefs and SE Ranking because their docs do too.

Run a new check, or read stored data?

A server that runs a new check sends your exact question to the AI engines now and returns the answers, brand mentions and cited sources. Use it for a new product, a client audit or a prompt nobody is tracking yet. Cite42 works this way, and you pay for each check.

A server that reads stored data gives the assistant answers a platform has already collected: Ahrefs' prompt index, SE Ranking's prompt database, or the daily runs in your OtterlyAI, Peec AI or LLM Pulse project. Use it for trends and share of voice across many prompts. The data is as fresh as the platform's last run and covers the prompts the platform already tracks, though several of these servers can add prompts for the platform to monitor from then on.

An SEO MCP server may also leave AI visibility out entirely. Semrush's server gives Claude its keyword, backlink and Trends data, while its AI Visibility Toolkit stays in the Semrush app (details below).

Cite42: new checks on any prompt, billed per call

Cite42 is a pay-per-call MCP server and API for AI visibility. Ask Claude something like "Check whether ChatGPT and Perplexity mention Acme, Beta or Gamma for best invoicing tool for freelancers, and list the sources they cite." The assistant calls Cite42's tools, which send that prompt to the AI engines you pick and return each answer's brand mentions, positions and cited URLs.

The tools an assistant gets:

• cite42_search: the full answer and sources from each AI engine.
• cite42_rankings: mention rate and average first-mention position for each brand, broken down by AI engine.
• cite42_citations: every cited source URL, and whether a page you're watching appears.
• cite42_compare and cite42_sentiment: head-to-head comparisons and how answers describe a brand.
• Tracker tools: save up to 30 prompts and rerun them weekly or monthly with stored history.

The same server covers keyword volume, Google Trends and content research workflows, so one conversation can go from "who gets recommended?" to "what should we write?".

There is no subscription: you top up a prepaid balance (minimum $25.00), the balance never expires, and you pay only for successful calls. A verified signup includes $1.00 of free credit. Each AI engine is priced separately, from $0.01. A rankings or citations check costs $0.05 on ChatGPT, $0.08 on ChatGPT and Perplexity, and $0.20 across all five, and the assistant can call cite42_pricing for free to quote a cost before it runs anything.

The hosted server at https://mcp.cite42.dev signs in with OAuth from Claude, Claude Code, ChatGPT (developer mode), Codex and Cursor. In Claude Code, add it with one command and sign in with claude mcp login cite42:

claude mcp add --transport http --scope user cite42 https://mcp.cite42.dev

The Claude Code and Codex guide covers sign-in and the other clients, and the ChatGPT guide covers developer mode. For setups that use an API key instead of OAuth, there's the npm package @cite42/mcp.

Pros

• Runs your exact prompt today on five AI engines, including Claude.
• No subscription or monthly minimum; a check across all five engines costs $0.20.
• The same tools are a REST API, so a check that works in Claude can move into your code unchanged.

Cons

• No database of past answers: you see only the prompts you or your trackers run.
• Trackers run weekly or monthly, and Copilot, Google AI Mode and Meta AI are outside its coverage.

The tracking guide shows a full 11-prompt baseline run with Cite42.

Ahrefs MCP: Brand Radar data in your assistant

Ahrefs' remote MCP server includes Brand Radar, its AI visibility tool. The tool categories page lists two Brand Radar groups, domain- and URL-level reports (brand-radar-simple) and entity-level reports (brand-radar-entities), plus a management group that covers Brand Radar configuration. Ahrefs' Brand Radar help article confirms the data can be pulled through the MCP server or the API.

Brand Radar's index holds more than 400 million search-backed prompts across Google AI Overviews and AI Mode, ChatGPT, Perplexity, Gemini and Copilot, with Claude tracked on custom prompts. The AI chatbot index refreshes monthly; custom prompts run on the schedule you set, from daily to monthly.

In Claude Code, add the server with claude mcp add ahrefs https://api.ahrefs.com/mcp/mcp -t http, then sign in through /mcp (setup guide). Ahrefs advertises more tools than some clients accept, so you can limit the connection to Brand Radar by adding ?tools=brand-radar-simple,brand-radar-entities to the URL.

The MCP server is available on paid plans from Lite. Each call spends API units from your plan's monthly allowance, at least 50 units per call, and Lite includes 200,000 units a month (getting started). Brand Radar's AI indexes are add-ons at $199 a month per platform or $699 a month for all platforms, and custom prompts are included from Lite (as of October 2026). Ahrefs' terms limit the MCP endpoint to AI assistants; scripts should use the Ahrefs API.

Pros

• Look up any brand or competitor in a large prompt index before writing prompts of your own.
• One connection also covers Ahrefs' backlink, keyword and rank tracking data.

Cons

• Needs an Ahrefs plan plus Brand Radar add-ons for the AI indexes, and every call spends API units.
• Chatbot data in the index updates monthly.

SE Ranking MCP: AI search data beside your SEO data

SE Ranking's hosted MCP server at https://api.seranking.com/mcp exposes more than 160 tools. Alongside keyword research, backlinks and site audits, its AI Search tools report brand mentions, visibility and source citations in ChatGPT, Perplexity, Gemini, Google AI Overviews and AI Mode (MCP guide). It signs in with OAuth or an API key. Sub-accounts can't use the sign-in button and need an API key from the account administrator.

The AI data comes from SE Ranking's AI Visibility API, built on more than 25 million prompts refreshed monthly, plus daily tracking for prompts you choose. Billing uses pay-as-you-go API credits: the API access page lists $50 for 250,000 credits, and each operation uses a different number of credits.

Pros

• AI search data and classic SEO data through one connection.
• Pay-as-you-go credits instead of a dashboard subscription.

Cons

• The prompt database refreshes monthly, and credit costs vary by operation.
• Claude is outside its AI coverage.

OtterlyAI: assistant access to a daily monitoring program

OtterlyAI's MCP server at https://data.otterly.ai/mcp connects an assistant to your OtterlyAI workspace. Read tools cover brand reports, citations, recommendations, audits, agent analytics and query fan-outs. Write tools can create prompts and start monitoring them "exactly like creation in the UI", and manage brand reports, tags and annotations (MCP reference). Sign-in is OAuth only; API keys are for the REST API.

MCP access starts on the Standard plan: $189 a month for 100 prompts tracked daily, with 2,000 MCP requests a month. Premium is $489 for 400 prompts and 5,000 requests (pricing, as of October 2026). Every tool call except get_account_info counts toward the limit. ChatGPT, Google AI Overviews, Perplexity and Copilot are included, and Google AI Mode, Gemini and Claude are paid add-ons.

Pros

• Daily monitoring plus an assistant that can read it and add prompts.
• Copilot is included on every plan.

Cons

• The $29 Lite plan has no MCP access.
• Every tool call counts toward the request limit, so long agent sessions use it up faster.

Peec AI: reports and prompt management from Claude

Peec AI's MCP server at https://api.peec.ai/mcp connects Claude, Cursor and other clients to your Peec AI data, signing in with OAuth or a personal access token (MCP docs). Read-only tools cover reports, chat inspection, source content and suggested actions. A smaller set of write tools creates prompts and manages topics, tracked brands and domain classifications, one at a time or in batches of up to 50. The documented engines are ChatGPT, Perplexity, Gemini, Google AI Overviews, Google AI Mode, Claude, Microsoft Copilot and Grok, limited in practice to the models on your plan.

In Claude Code, the command is claude mcp add peec-ai --transport http https://api.peec.ai/mcp. Peec AI's MCP docs leave plan eligibility unstated, and its REST API is limited to Enterprise customers, so confirm MCP access with Peec AI before you buy.

Pros

• Read and write tools, including bulk prompt creation.
• Unlimited users on every Peec AI plan.

Cons

• Which plans include MCP is undocumented.
• Self-serve plans track three models, so the assistant sees only those.

LLM Pulse, Profound and Scrunch: other monitoring platforms

LLM Pulse includes its MCP server, at https://api.llmpulse.ai/api/v1/mcp, on every plan. Its roughly 90 tools cover visibility metrics, share of voice, mentions, citations, sentiment, prompts and responses, and the assistant can create prompts, add competitors and launch recommendation runs (MCP guide). Starter costs €49 a month for 50 prompts tracked weekly on ChatGPT, Perplexity, Gemini, Google AI Mode and AI Overviews (as of October 2026).

Profound runs a hosted MCP server at https://mcp.tryprofound.com/mcp for visibility, sentiment, citation and shopping reports, agent analytics and prompt management. It's available on the Enterprise plan, which is priced by quote.

Scrunch offers an MCP connection for querying your monitoring data from Claude, ChatGPT, Copilot, Cursor and other assistants, and it needs a Scrunch account with active monitoring (help center). Scrunch plans start at $250 a month billed annually. The AI visibility tools guide compares these dashboards in more detail.

Semrush, DataForSEO and community servers

Two names from SEO MCP lists are left off this shortlist. Semrush's MCP server at https://mcp.semrush.com/v2/mcp gives assistants its SEO, Trends and read-only Projects data, and its developer docs list no AI Visibility Toolkit or Brand Performance tools as of October 2026. DataForSEO's official MCP server is a general tool that browses its API documentation and sends authenticated requests; its LLM Mentions database is covered in the API comparison. The SEO MCP comparison covers both as SEO data servers.

Directories such as mcp.so and Glama also list dozens of community AI visibility servers on GitHub and npm. Some wrap DataForSEO's API, and others call the AI models directly with keys you supply. Check who maintains a server, which keys it needs and where it sends your prompts before you connect it.

How to test a server before you rely on it

1. Pick three prompts your buyers ask and three to five brands, including yours.
2. Ask the assistant to list the server's tools and say whether each one runs a new check or reads stored data.
3. Run one prompt and read the raw answer and cited sources as well as any score.
4. Check what the session cost: per-call charges, API units or the share of your monthly request allowance.
5. Decide whether you need new checks, stored history or both. Plenty of teams use a dashboard for daily tracking and a pay-per-call server for one-off questions.

Frequently asked questions

Which MCP servers can check my brand's visibility in ChatGPT?

Cite42, Ahrefs, SE Ranking, OtterlyAI, Peec AI, LLM Pulse and Profound all offer MCP servers with ChatGPT visibility data. Cite42 runs a new ChatGPT check on any prompt for $0.05. The others read data their platforms collected: Ahrefs from Brand Radar's prompt index, SE Ranking from its prompt database, and OtterlyAI, Peec AI, LLM Pulse and Profound from the prompts you track with them.

Is there an AI visibility MCP server without a subscription?

Yes. Cite42's MCP server bills per successful call with no subscription: you top up from $25.00, the balance never expires, and a brand check across ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews costs $0.20. SE Ranking's MCP server runs on its API access, which is also sold as pay-as-you-go credits, from $50 for 250,000.

Does the Ahrefs MCP server include Brand Radar data?

Yes. Ahrefs' MCP documentation lists Brand Radar tool groups for domain-, URL- and entity-level reports, and the server is available on Ahrefs plans from Lite. The AI chatbot and AI Overview indexes need Brand Radar add-ons ($199 a month per platform or $699 for all, as of October 2026), and every call uses API units.

Does the Semrush MCP server include AI Visibility Toolkit data?

Semrush's developer documentation lists its SEO, Trends and read-only Projects APIs for the MCP server, with no AI Visibility Toolkit tools, as of October 2026. Use the toolkit in the Semrush app, or connect one of the servers in this guide if you need AI visibility data inside Claude or ChatGPT.

What's the difference between an SEO MCP server and an AI visibility MCP server?

An SEO MCP server returns search data such as keyword volume, rankings and backlinks. An AI visibility MCP server reports what AI assistants say: which brands an answer mentions, in what order, and which sources it cites. Ahrefs, SE Ranking and Cite42 offer both through one connection, and the SEO MCP comparison covers the SEO side.

Connect your first server

Choose by the job. Cite42 fits fresh checks on your own prompts without a subscription, Ahrefs fits competitor research in a large prompt index, and OtterlyAI, Peec AI, LLM Pulse and Profound fit teams that already run a monitoring program and want it inside Claude or ChatGPT. To try Cite42, create a free account for $1.00 of starter credit, connect https://mcp.cite42.dev in your assistant and ask about one buyer prompt. The docs list every tool, and the API comparison covers the same checks over REST.