MCP servers: what they are and how to use them

An MCP server connects your AI assistant to external tools and live data. Instead of copying information between apps, your AI calls the server directly and gets structured results back in your conversation. Trends MCP is an MCP server for live trend data across Google, TikTok, YouTube, Reddit, Amazon, and 10+ more sources.

get_trends

Query historical trend data for any keyword across any of the 15+ supported sources - returns a normalized time series your AI can analyze and narrate.

get_trends(keyword='mcp server', source='google search', data_mode='weekly')

get_growth

Measure how fast any topic is growing across platforms - ask your AI 'Is interest in MCP servers growing on Google and Reddit?' and get a structured data-backed answer.

get_growth(keyword='mcp server', source='google search, reddit', percent_growth=['3M', '1Y'])

get_ranked_trends

Surface the fastest-growing keywords or topics right now from any data source - useful for trend discovery without knowing what to look for in advance.

get_ranked_trends(source='google search', sort='wow_pct_change', limit=20)

get_top_trends

Get live trending feeds from TikTok, Google, Reddit, or any supported platform - no keyword needed, surfaces what is breaking right now.

get_top_trends(type='TikTok Trending Hashtags', limit=20)

Common questions

An MCP server is a program that exposes tools and data to AI assistants using the Model Context Protocol (MCP), an open standard created by Anthropic. When you connect an MCP server to your AI, the AI can call the server's tools directly during a conversation - retrieving live data, running queries, or taking actions - without you having to copy and paste information between apps.
Your AI client (Claude, Cursor, VS Code, etc.) connects to the MCP server over HTTP or a local process. When you ask your AI a question that requires external data, it calls the relevant MCP tool, the server returns a structured JSON response, and your AI incorporates that data into its answer. The whole exchange happens inside your existing conversation.
The most widely used MCP servers cover: live data (Trends MCP for search, social, and web trends), file systems (built-in filesystem server), databases (Postgres and SQLite MCP servers), web browsing (Playwright and Puppeteer MCP), code execution (Python and shell execution servers), and productivity tools (Notion, Linear, GitHub, Slack MCP servers). The best server depends on what data or capability your AI workflow needs.
Claude Desktop, Claude.ai (Pro, Max, Team), Cursor, VS Code with GitHub Copilot, Windsurf, Cline, Raycast, ChatGPT (via custom integrations), and OpenAI-compatible clients. MCP is an open protocol so adoption is growing rapidly across AI tools.
For most clients: add a JSON config snippet to your MCP settings file specifying the server URL and your API key. For Claude.ai in the browser: go to Settings → Connectors → Add custom connector. For Cursor: edit ~/.cursor/mcp.json. Trends MCP has one-click install buttons for every supported client.
Trends MCP is an MCP server that gives your AI live trend data. It exposes tools like get_trends, get_growth, get_ranked_trends, and get_top_trends - covering Google Search, YouTube, TikTok, Reddit, Amazon, Wikipedia, news, web traffic, app downloads, Steam, and npm. One connection gives your AI access to all 15+ sources with no separate API keys.
Get MCP servers: what they are and how to use them in 30 seconds
Free tier includes 100 requests per month. No credit card required. Works with Claude, Cursor, ChatGPT, Raycast, and every MCP client.
Get your free API key