Manifold MCP icon

Manifold MCP

Manifold MCP is a hosted remote MCP server that lets Claude, ChatGPT, Cursor, and other MCP clients access SEO, analytics, AI-visibility, social, advertising, and lead-research data. It is intended for read-only marketing research, with most calls charged against workspace credits and 250 credits available to start.

Manifold MCP

Marketing research tools for AI agents

Manifold MCP is a hosted marketing MCP server that gives Claude, ChatGPT, Cursor, and other MCP clients access to SEO, analytics, AI-visibility, social, advertising, and lead-research tools. After one connection, an AI agent can choose the relevant lookup tools from a single remote server rather than requiring a separate server for each data source.

The service is designed for research and analysis rather than publishing or account modification. It retrieves Google search and first-party site data, public social and advertising information, AI-generated answer observations, and people or company data. Usage is charged from workspace credits for most calls, with controls for estimating and limiting call cost before work begins.

Key features

SEO and search research

Search Google organic results for keywords, rankings, result pages, domains, backlinks, and site health, with parameters such as location, language, device, and result limit.

First-party site analytics

Connect data from Google Search Console, Bing Webmaster Tools, and Google Analytics 4 to inspect clicks, rankings, indexing, crawl issues, sessions, conversions, revenue, and campaign or channel performance. These tools are listed as free.

AI visibility analysis

Query what ChatGPT, Claude, Gemini, Perplexity, and Google’s AI answers say and cite, and assess whether a site is ready for AI answer experiences.

Advertising and lead research

Search public ad libraries and investigate people, companies, and verified work emails. The documentation notes that every email reveal is charged.

Social and community data

Look up content and audience signals across Reddit, TikTok, Instagram, Facebook, YouTube, LinkedIn, and X, including platform-specific posts, comments, accounts, transcripts, pages, channels, or company information.

Agent workflow controls

Use one hosted remote MCP server with agent-selected tools, credit estimates, caching, pagination, asynchronous tasks, and metadata describing each response.

Practical use cases

  • SEO research and site audits

    Ask an agent to investigate a keyword, inspect Google result pages, compare ranking domains, review backlinks, or run a technical SEO audit. The agent can combine search data with a target domain or URL.

  • AI visibility monitoring

    Review what major AI systems say about a brand or topic, which sources they cite, and whether a site is prepared for AI answer surfaces. The AI-answer operation can run as an asynchronous task for longer work.

  • Performance and measurement analysis

    Combine Search Console, Bing Webmaster Tools, and Google Analytics 4 data to examine search performance, indexing and crawl issues, traffic, conversions, revenue, channels, pages, or campaigns for connected sites.

  • Social and audience research

    Search Reddit communities and other social platforms for posts, comments, videos, transcripts, accounts, or audience signals to inform market and community research.

  • Lead and competitor research

    Find companies and people and verify work emails, or inspect public ad libraries by keyword, brand, or advertiser. Email reveals consume credits.

Pros and Cons

Pros

  • One hosted MCP connection covers multiple marketing data groups, including SEO, AI visibility, social, ads, analytics, and lead research.
  • Supports Claude, ChatGPT, Cursor, and other MCP clients, with browser-based sign-in or API-key access for scripts.
  • Provides cost visibility and safeguards through dry_run estimates, max_credits limits, per-result credit metadata, and caching.
  • Includes pagination, batch-input handling, hints for next tool calls, and task polling for longer operations.
  • A free workspace starts with 250 credits, and Google Search Console and Google Analytics tools are listed as free.

Cons

  • Most lookups use credits, and lead-data email reveals are charged; users need to manage usage for repeated or broad research.
  • The tools are read-only: the documentation says they do not post, send messages, or edit anything.
  • Some operations, including technical crawls and AI-answer collection, are asynchronous rather than returning immediately.

FAQ

How do I connect Manifold MCP to an AI client?

Add the hosted server URL, https://mcp.manifoldmcp.com/mcp, to a compatible client or run the client’s setup command. The client opens a browser for sign-in; users can authenticate with Google or email and approve Manifold. The docs state that setup works with Claude, ChatGPT, Cursor, and other MCP clients.

Can Manifold MCP publish content or change data?

No. Manifold’s tools look up data and do not post, send messages, or edit anything. Results include call metadata such as the provider, data date, rows returned, and credits charged.

How does usage and credit consumption work?

Calls use credits from a workspace balance. Most calls cost credits, while Google Search Console and Google Analytics tools are described as free. You can use dry_run to get an estimate without charge and max_credits to refuse a call above a chosen limit. Cached answers cost 25% of the live price, or nothing when the workspace has already paid for that result.

Is there a free tier or paid pricing?

The free workspace includes 250 credits with no card required and no renewal. The pricing page also lists Pro at $20 per month with 2,500 credits and Max at $250 per month with 42,000 credits; credits can also be topped up. Plan credits renew monthly, while top-up credits do not expire.

What do tool results and longer-running jobs include?

Manifold returns structured data together with metadata and hints for possible next calls. List tools support pagination with an opaque cursor, and two longer-running operations—technical SEO crawling and AI-answer collection—run as asynchronous tasks that can be checked with get_task.

Quick Facts

Category
Developer tool; SEO and marketing research
Platform
Hosted remote MCP server
Compatible clients
Claude, ChatGPT, Cursor, and other MCP clients
Primary data areas
SEO, Search Console, Analytics, AI visibility, social, ad libraries, and people/company data
Free access
250 workspace credits; no card required
Pricing
Pro from $20/month; Max from $250/month; top-ups available

Alternativas a Manifold MCP