npm.io
1.0.6 • Published 2d ago

n8n-nodes-exa-unofficial

Licence
MIT
Version
1.0.6
Deps
0
Size
293 kB
Vulns
0
Weekly
0

n8n-nodes-exa-unofficial

This unofficial n8n community node provides Exa search, content extraction, answers, Agent runs, and Websets.

Exa is an AI-powered search engine that provides embeddings-based semantic search, deep web research, and structured content extraction.

n8n is a fair-code licensed workflow automation platform.

Difference from the official Exa node
Installation
Operations
Credentials
Compatibility
Usage
Resources

Difference from the official Exa node

The official Exa community node covers the core Search, Deep Search, Contents, and Answer APIs. This unofficial node includes those operations and additionally exposes Exa's Agent and Websets APIs directly in n8n.

Extra resource Options provided by this node
Agent Create, get, list, cancel, and delete runs; list run events; select effort; continue from a previous run; set a system prompt; pass JSON input, output schemas, and metadata; use additional data providers such as Similarweb, Financial Datasets, and Particle.
Websets Create and preview Websets; get, list, update, cancel, and delete them; manage searches and items; create and manage enrichments; create CSV or JSON exports and check export status.

Use the official node when Search, Contents, or Answer is sufficient. Use this node when a workflow also needs long-running Agent tasks, structured Webset collection building, enrichment, or export management without assembling raw HTTP Request nodes.

Installation

Follow the installation guide in the n8n community nodes documentation.

Operations

This node supports the following Exa API operations:

  • Search: Intelligently search the web
    • Supports multiple search types: auto, fast, instant
    • Filter by domain, date, category, and text content
    • Extract page contents (text, highlights, summaries)
    • Configure livecrawl and subpage crawling
Contents
  • Get Contents: Retrieve cleaned text, highlights, and summaries from a list of URLs
Answer
  • Get Answer: Get an AI-generated web-grounded answer to a query
Agent
  • Create, Get, List, Cancel, Delete Run: Run Exa Agent tasks and manage lifecycle.
  • List Run Events: Fetch JSON event history with cursor pagination. SSE not exposed because standard n8n node execution expects completed JSON response.
Websets
  • Websets: Create, preview, get, list, update, cancel, and delete Websets.
  • Searches: Create, get, and cancel Webset searches.
  • Items: List, get, and delete Webset items.
  • Enrichments: Create, get, update, cancel, and delete enrichments.
  • Exports: Create CSV or JSON exports and poll export status.

Credentials

To use this node, you need an Exa API key. You can obtain one by:

  1. Signing up at dashboard.exa.ai
  2. Navigating to the API Keys section
  3. Creating a new API key

Add the API key to your n8n credentials as "Exa (Unofficial) API".

Compatibility

  • Minimum n8n version: 1.0.0
  • Tested against n8n version: 1.60.0

Usage

Basic Search Example
  1. Add the Exa (Unofficial) node to your workflow
  2. Select "Search" as the resource
  3. Enter your search query (e.g., "Latest AI research papers")
  4. Configure search type (auto, fast, or instant)
  5. Optionally add filters:
    • Category (research paper, news, company, etc.)
    • Date range (published date)
    • Domain filters (include/exclude specific domains)
    • Text filters (include/exclude specific text)
Content Extraction

Enable content extraction options:

  • Text: Get cleaned page text
  • Highlights: Get relevant excerpts
  • Summary: Get AI-generated summaries
  • Livecrawl: Crawl pages in real-time
  • Subpages: Crawl linked subpages
Advanced Features

Date Filtering: Filter results by publication date

Start Published Date: 2024-01-01
End Published Date: 2024-12-31

Domain Control: Focus on or exclude specific domains

Include Domains: arxiv.org, github.com
Exclude Domains: reddit.com

Category Filtering: Focus on specific content types

  • Company profiles
  • Research papers
  • News articles
  • PDFs
  • GitHub repositories
  • Twitter/LinkedIn profiles
  • Financial reports
Agent Workflow Example
  1. Set Resource to Agent, then select Create Run.
  2. Enter query, optional system prompt, JSON input/output schema, metadata, and data sources.
  3. Store returned run id, then use Get Run with {{$json.id}} to poll until complete. No automatic polling occurs.
Websets Example
  1. Set Resource to Websets, then select Create Webset.
  2. Enter query and count; optionally provide entity, criteria JSON, enrichment JSON, title, external ID, and metadata.
  3. Use returned Webset ID in Get Webset, List Items, Create Enrichment, or Create Export through n8n expressions.

For bulk monitoring, use List Research Tasks and apply parameters for pagination, status filtering, or creation date ranges.

Resources

Development

To work on this node locally:

# Install dependencies
npm install

# Build the node
npm run build

# Link for local testing
npm link

# In your n8n installation directory
cd ~/.n8n/custom
npm link n8n-nodes-exa-unofficial

# Start n8n
n8n start

License

MIT

Keywords