Fast News MCP Server

Real-time news search, trending headlines, topic categories, and local news — exposed as MCP tools so your AI agent can query news in plain language. No HTTP boilerplate required.

The server is hosted by RapidAPI and connects via mcp-remote. All you need is a RapidAPI subscription and your API key.

Demo

See the Fast News API in action: the-dispatcher-demo.vercel.app

Get an API Key

1. Subscribe on RapidAPI (free tier available)
2. Your API key appears on the Endpoints tab

Every request needs two headers:

| Header | Value | |--------|-------| | x-rapidapi-key | Your API key | | x-rapidapi-host | fast-news-with-previews.p.rapidapi.com |

Quick Start

Prerequisites: Node.js (for npx) and a free RapidAPI subscription.

Add the following to your AI agent's MCP server configuration. Replace YOUR-API-KEY with your RapidAPI key.

{
  "mcpServers": {
    "fast-news": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.rapidapi.com",
        "--header",
        "x-api-host: fast-news-with-previews.p.rapidapi.com",
        "--header",
        "x-api-key: YOUR-API-KEY"
      ]
    }
  }
}

The mcp-remote package is installed automatically by npx. You don't need to install anything manually.

See the per-agent setup instructions below for client-specific details.

Prefer the Quick Start above. Use this only if your setup specifically benefits from a single published package.

This repository is also published as the npm package @wllrdev/fast-news-mcp. It's a thin wrapper around mcp-remote that reads your key from the RAPIDAPI_KEY environment variable instead of the command line:

{
  "mcpServers": {
    "fast-news": {
      "command": "npx",
      "args": ["-y", "@wllrdev/fast-news-mcp"],
      "env": {
        "RAPIDAPI_KEY": "YOUR-API-KEY"
      }
    }
  }
}

Available Tools

Once connected, your AI agent has access to these tools:

| Tool | Description | |------|-------------| | search_news | Search any topic with filters for language, country, and timeframe. Supports source diversity and built-in deduplication. | | advanced_search | Like search_news but with post-fetch controls — include/exclude specific sources, require thumbnails, and change sort order. | | trending_headlines | Top headlines auto-translated into 40+ languages. Optionally filter by country and timeframe. | | topic_categories | Browse 11 predefined categories: world, national, business, technology, entertainment, sports, science, health, politics, economy, environment. Combine two categories in one request. | | local_news | News about a specific city, region, or area. |

For full parameter documentation, see the API reference.

Setup for Your AI Agent

Find your client below and expand it for setup steps.

Option A — CLI command:

claude mcp add fast-news -- npx mcp-remote https://mcp.rapidapi.com --header "x-api-host: fast-news-with-previews.p.rapidapi.com" --header "x-api-key: YOUR-API-KEY"

Add --scope user to make it available globally across all projects.

Option B — Manual config:

Add the Quick Start config to your Claude Code MCP settings.

1. Open Claude Desktop settings
2. Navigate to Developer > Edit Config
3. Paste the Quick Start config into claude_desktop_config.json
4. Restart Claude Desktop

1. Open Cursor Settings > MCP > New MCP Server
2. Paste the Quick Start config
3. Replace YOUR-API-KEY with your key
4. Enable the server

Add the following to your .vscode/mcp.json:

{
  "servers": {
    "fast-news": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.rapidapi.com",
        "--header",
        "x-api-host: fast-news-with-previews.p.rapidapi.com",
        "--header",
        "x-api-key: YOUR-API-KEY"
      ]
    }
  }
}

Add the Quick Start config to your Windsurf MCP config file (~/.codeium/windsurf/mcp_config.json).

1. Open Cline settings
2. Go to MCP > Add New MCP Server
3. Select Local and paste the Quick Start config

Add the Quick Start config to your ~/.continue/config.json under the mcpServers key.

Add the following to your opencode.json:

{
  "mcp": {
    "fast-news": {
      "type": "remote",
      "url": "https://mcp.rapidapi.com",
      "headers": {
        "x-api-host": "fast-news-with-previews.p.rapidapi.com",
        "x-api-key": "YOUR-API-KEY"
      }
    }
  }
}

codex mcp add fast-news -- npx mcp-remote https://mcp.rapidapi.com --header "x-api-host: fast-news-with-previews.p.rapidapi.com" --header "x-api-key: YOUR-API-KEY"

Usage Examples

Once connected, try these prompts in your AI agent:

• "What are the trending headlines right now?"
• "Search for news about artificial intelligence"
• "Find local news about Tokyo"
• "Show me the latest technology news from the past week"
• "Search for climate news from BBC and Reuters only, sorted by newest first"
• "Get trending headlines in German"
• "Find sports and business news combined"
• "Show me health news with thumbnails only from the last 7 days"

The agent will use the MCP server to call the appropriate endpoint and return formatted results.

API Documentation

For full endpoint documentation, parameters, response formats, and code examples:

• Overview: docs.marvelousapi.com/apis/fastnews
• Endpoints: docs.marvelousapi.com/apis/fastnews/docs/endpoints/news
• Error Handling: docs.marvelousapi.com/apis/fastnews/docs/error-handling

Key Features

• Source diversity — no single domain dominates your feed (max 25% per domain)
• Built-in deduplication — near-identical articles are collapsed automatically
• 40+ languages — trending and topic searches auto-translate into local queries
• Clean previews — up to 250-character excerpts with thumbnail URLs
• Timeframe filters — last hour, 24h, 7 days, 30 days

Links

• Subscribe on RapidAPI
• API Documentation
• Live Demo
• Discord

License & API Terms

The configuration and documentation in this repository are MIT licensed. The Fast News API itself is a commercial service hosted on RapidAPI — usage is subject to the RapidAPI Terms of Service and your subscription plan. The MIT license does not grant access to the API.

License

MIT