Give your agents Nowing as native tools
The Nowing MCP server hands Claude, Cursor, or any MCP client the whole platform: scrape Reddit, YouTube, Instagram, TikTok, Amazon, Google Maps, Google Search, and the open web, and search, read, and write your knowledge base. One API key, typed tools, pay as you go.
.cursor/mcp.json
{
"mcpServers": {
"nowing": {
"url": "https://mcp.nowing.com/mcp",
"headers": {
"Authorization": "Bearer nw_pat_..."
}
}
}
}Works with Claude Code, Cursor, Claude Desktop, and any MCP client
From API key to agent tools in three steps
Create an API key
In Nowing, go to Settings, then API, and create a key. Enable API access on the workspaces you want your agents to reach. That key is all the server needs.
Add the server to your client
Point your client at https://mcp.nowing.com/mcp with your key in an Authorization header — the hosted config for Cursor, Claude Code, and others is one paste. Prefer stdio? Switch to Self-host and run it against your own backend.
Your agent has the tools
Every scraper and knowledge base operation shows up as a native, typed MCP tool. Your agent picks a workspace once and the server carries the context between calls.
Step-by-step setup for every agent
Pick your client, choose Hosted or Self-host, and paste the config. Replace the key with one from API Playground → API Keys — or grab a pre-filled config from the playground itself.
- Run this command in a terminal (any directory).
- Start Claude Code and run /mcp — nowing should be listed as connected.
Terminal
claude mcp add --transport http nowing https://mcp.nowing.com/mcp \
--header "Authorization: Bearer nw_pat_your_key_here"Every tool the server exposes
The server is a thin layer over the Nowing REST API: the same endpoints, the same billing, no backend code imported. Whatever ships in the API shows up here.
Live scrapers
Structured, current platform data. One returned item is one billable unit.
- nowing_reddit_scrape
- nowing_youtube_scrape
- nowing_youtube_comments
- nowing_instagram_scrape
- nowing_instagram_details
- nowing_tiktok_scrape
- nowing_tiktok_comments
- nowing_tiktok_user_search
- nowing_tiktok_trending
- nowing_google_maps_scrape
- nowing_google_maps_reviews
- nowing_google_search
- nowing_amazon_scrape
- nowing_web_crawl
- nowing_list_scraper_runs
- nowing_get_scraper_run
Knowledge base
Read and write the same knowledge base your Nowing agents use.
- nowing_search_knowledge_base
- nowing_list_documents
- nowing_get_document
- nowing_add_document
- nowing_upload_file
- nowing_update_document
- nowing_delete_document
Workspace selector
Pick a workspace once; every later call defaults to it.
- nowing_list_workspaces
- nowing_select_workspace
The Nowing MCP server vs external MCP connectors
They are two sides of the same protocol. The MCP server on this page pushes Nowing tools out to agents you already run in Claude, Cursor, or your own harness. External MCP connectors do the reverse: they pull outside tools like Notion, Slack, and Jira into your Nowing agents. Use both and data flows in either direction.
Nowing MCP server: frequently asked questions
Put live web data inside your agents
The MCP server is part of the Nowing open web research platform. Start with $5 of free credit, no credit card required.
NowingNowing