Grok Collections Upload

Upload documents to Grok collections for RAG.

Overview

Grok Collections Upload sends binary file data to an xAI Grok collection for retrieval-augmented generation. Takes binary input (PDF, text, and so on) from an upstream node and uploads it as a document to the specified collection. Returns the document metadata including the document ID.

Category: AI
Tool Name: grok_collections_upload
Version: 1

Appearance: Icon: brain | Color: #000000

Node Type

Action (Binary) — handles file/binary data operations

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
Collection ID	string	Yes	—	The ID of the collection to upload to. Falls back to the input item’s “collectionId” field. Supports expressions.
Binary Property Name	string	No	data	Name of the binary property containing the file to upload. Names are case-sensitive — see the upstream node’s Binary Data panel for the exact names to use.
Options	collection	No	{}	Optional output settings — add only the fields you want to override.
— Response Field Name	string	No	document	The output field name where the upload response will be stored.
Include Input	boolean	No	false	Whether to include the original input item fields in the output.
Max Concurrency	number	No	5	Maximum number of items to process concurrently. Lower default to avoid overwhelming the upload endpoint.

Output Data

One output item per input item. The xAI response is written unchanged to the field named by Response Field Name (document by default), with two convenience fields alongside it. The rest of the input JSON is carried over only when Include Input is on; the input item’s binary data is forwarded untouched.

• document — the upload response from xAI, under whatever name Response Field Name is set to.
• collectionId — the collection the file was uploaded to, as resolved for this item.
• documentId — the ID xAI assigned to the uploaded document.

The file itself is read from the named binary property. If that property is missing on an item, the item fails.

Reference the result downstream by expression, e.g. {{ $json.documentId }}.

Usage Examples

• Upload a PDF to a Grok collection for RAG
• Add text files to a knowledge base collection
• Upload documents from upstream binary data
• Batch upload multiple files to a collection
• Import documents for Grok retrieval-augmented generation

Example Configuration

Upload a downloaded file to a fixed collection:

{
  "type": "grok_collections_upload",
  "parameters": {
    "collectionId": "{{ $json.collectionId }}",
    "binaryPropertyName": "data"
  }
}

Upload from a named binary property and keep the source metadata:

{
  "type": "grok_collections_upload",
  "parameters": {
    "collectionId": "{{ $json.collectionId }}",
    "binaryPropertyName": "pdfFile",
    "includeInput": true,
    "maxConcurrency": 3,
    "options": {
      "responseFieldName": "uploadResult"
    }
  }
}

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 Collections Upload sends binary file data such as PDFs or text files to an xAI Grok collection via a multipart POST request for Retrieval-Augmented Generation. Use this tool when you need to populate a Grok RAG knowledge base with documents so that future queries can retrieve and reference their content. It returns document metadata on the main output including the assigned document ID, or routes failures to the error output.