Give your n8n AI agent Google search and live data

Set up the MCP Client Tool node

1. Create an API key on the Keys page of the 1ToolRing console. It is shown once, so copy it now.
2. Add an AI Agent node with a chat model, then add an MCP Client Tool node to its Tool connector.
3. In the MCP Client Tool node, set Endpoint to the server URL:
   Copy

   https://mcp.1toolring.com/mcp

4. Set Server Transport to HTTP Streamable.
5. Set Authentication to Bearer Auth, create a new credential and paste your API key as the token.
6. Leave Tools to Include on All. The server lists exactly four tools.
7. Open the chat, send the first prompt below and watch the agent call discover_tools, inspect_tool and run_tool.

Without an agent: the HTTP Request node

When the workflow always runs the same tool (say, a daily rank check for a list of keywords), skip the agent and call the REST API at https://api.1toolring.com directly. Find the tool id once, with discover_tools in a chat or with this call:

curl -X POST https://api.1toolring.com/api/v1/tools/discover \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"google search results for a keyword","limit":3}'

Configure the HTTP Request node

Method POST, URL https://api.1toolring.com/api/v1/tools/run. For Authentication choose Generic Credential Type, then Bearer Auth, with the same API key. Send a JSON body like this one, where {{ $json.keyword }} comes from the previous node:

{
  "tool": "dataforseo/serp/google-organic",
  "input": {
    "body": {
      "keyword": "{{ $json.keyword }}",
      "language_code": "id"
    }
  }
}

Slow runs and cost

A run that finishes in time answers 200 with the result in data.output and the charge in data.cost ({"usd", "idr"}). A slow tool answers 202 with a run_id: add a Wait node, then a GET request to the URL below, which is free, until status is completed or failed.

The X-ToolRing-Cost-USD and X-ToolRing-Cost-IDR response headers hold the exact charge if you want to log it. Read the input fields for any tool with POST /api/v1/tools/inspect; it is free too.

https://api.1toolring.com/api/v1/tools/runs/{{ $json.data.run_id }}

First prompt to try

Send this in the AI Agent chat. The agent should look up a tool, tell you the price, and then run it.

Use 1toolring to get the top Google results for "standing desk" in the United States. Tell me which tool you will use and its price before you run it, then list the top 10 URLs with their rank.

What the calls cost

A few tools your agent is likely to pick. Discovering and inspecting tools is free; only a successful run is charged. How pricing works.

Troubleshooting

Questions

Related

Tool pages

• Google search
• Keyword research
• Google Maps places
• Amazon products
• Exchange rates

Other guides

• Claude Code
• Cursor
• Codex
• VS Code / Copilot
• ChatGPT
• All clients on the install page

Steps checked against n8n docs: MCP Client Tool node (checked 2026-11-05).