Grok Web Search

Search the web using xAI Grok models with live search capabilities.

Overview

Grok Web Search asks a Grok model a question and lets it search the web while answering. The search runs on xAI’s servers; Grok decides when to use it and comes back with a synthesized answer plus the sources it cited. Search can be left to the model’s judgement, attached to every request, or switched off so the node behaves like ordinary chat. Returns the response text, an array of sources, model name, token usage, and status.

Category: AI
Tool Name: grok_web_search
Version: 1

Appearance: Icon: brain | Color: #000000

Node Type

Action — processes input items and produces output

Input / Output

Direction	Port(s)
Input	Input
Output	Output, Error

Credentials

This tool requires xAI credentials. See the Credentials Guide for setup instructions.

Parameters

Parameter	Type	Required	Default	Description
Model	options	No	Platform default	The Grok model to use for web search.
				Options: the Grok chat models currently available — the dropdown tracks the model catalog, so it changes as xAI’s line-up changes.
System Prompt	string	No	—	Optional system prompt to set the behavior and context for the model. Supports expressions like {{ $json.persona }}.
User Message	string	Yes	—	The search query or prompt to send to Grok. If empty, falls back to the input item’s “message” or “prompt” field. Supports expressions.
Search Mode	options	No	auto	Controls when the model performs a web search.
				Options: auto (model decides when to search the web based on the query), on (always perform a web search for every query), off (never search the web — behaves like regular chat)
Options	collection	No	{}	Optional generation and output settings — add only the fields you want to override.
— Max Tokens	number	No	4096	Maximum number of tokens to generate in the response.
— Temperature	number	No	1	Controls randomness. Lower values make output more focused; higher values more creative. Range: 0-2.
— Return Citations	boolean	No	true	Whether to include source citations in the response.
— Allowed Domains	string	No	—	Restrict search to these domains. Comma-separated. Cannot be combined with Excluded Domains — if both are set, Allowed Domains wins.
— Excluded Domains	string	No	—	Exclude these domains from search. Comma-separated. Ignored when Allowed Domains is set.
— Response Field Name	string	No	response	The output field name where the AI response text will be stored.
Include Input	boolean	No	false	Whether to include the original input item fields in the output alongside the AI response.
Max Concurrency	number	No	10	Maximum number of items to process concurrently.

Output Data

One output item per input item. The answer text is written to the field named by Response Field Name (response by default). The rest of the input JSON is carried over only when Include Input is on; binary data from the input item is forwarded.

• response — the synthesized answer, under whatever name Response Field Name is set to.
• sources — an array of the pages the answer cited, each { "url": "...", "title": "..." }. title is an empty string when the source did not supply one. The array is always present, and is empty when the model answered without citing anything (including every run with Search Mode set to off).
• model — the model that produced the answer.
• usage — the token counts reported by xAI for the request.
• finishReason — the status xAI reported for the request.

Reference the result downstream by expression, e.g. {{ $json.response }} or {{ $json.sources[0].url }}. To fan out one item per source, follow this node with a Split Out node on sources.

Usage Examples

• Search the web for current news using Grok
• Research a topic with live web search and citations
• Restrict a search to a trusted set of domains
• Get up-to-date answers from Grok with source citations
• Query the web using Grok with auto search mode

Example Configuration

Ask a question and let the model decide whether to search:

{
  "type": "grok_web_search",
  "parameters": {
    "userMessage": "What are the latest breakthroughs in quantum computing?"
  }
}

Force a search on every item and keep the answer short:

{
  "type": "grok_web_search",
  "parameters": {
    "userMessage": "{{ $json.query }}",
    "searchMode": "on",
    "maxConcurrency": 5,
    "options": {
      "maxTokens": 512,
      "temperature": 0.2,
      "responseFieldName": "answer"
    }
  }
}

Use the node as plain chat, with search disabled:

{
  "type": "grok_web_search",
  "parameters": {
    "systemPrompt": "You explain technical concepts simply.",
    "userMessage": "Explain how transformer attention mechanisms work.",
    "searchMode": "off",
    "options": {
      "maxTokens": 4096,
      "responseFieldName": "explanation"
    }
  }
}

Confine the search to a trusted allow-list (comma-separated, and mutually exclusive with Excluded Domains):

{
  "type": "grok_web_search",
  "parameters": {
    "userMessage": "Summarize major AI policy developments in the EU",
    "searchMode": "on",
    "options": {
      "allowedDomains": "europa.eu, consilium.europa.eu",
      "responseFieldName": "policyDigest"
    }
  }
}

Error Handling

Mode	Behavior
stop	Halts workflow on first error
continue	Skips failed items, passes successful ones through
errorPort	Routes failed items to Error output port

Tips

Grok Web Search queries the live internet using xAI Grok models with configurable search modes and optional domain allow/deny filtering to retrieve real-time synthesized answers. There is no date-range parameter — express recency in the user message and check the returned source URLs. Use it when your workflow requires current facts, breaking news, or time-sensitive data that a static model cannot accurately provide. It outputs a synthesized response text and an array of citation objects on the main channel, or an error object on the error channel.