Firecrawl API SOP

Appearance

Firecrawl + n8n Integration ​

Updated Feb 2026

Firecrawl combined with n8n enables web scraping automation through a visual workflow builder with 400+ integrations. No programming knowledge required -- build powerful scraping pipelines by connecting nodes on a canvas.

Account Setup ​

Firecrawl API Key ​

  1. Visit firecrawl.dev and create an account via email or GitHub
  2. Navigate to the API Keys dashboard
  3. Generate a new API key with a descriptive name (e.g., n8n Integration)
  4. Store the key securely -- treat it like a password

n8n Deployment Options ​

Cloud Version (Recommended for beginners):

  • Visit n8n.cloud and register
  • No installation required; managed infrastructure included
  • Free tier available for testing

Self-Hosted via Docker:

bash
docker volume create n8n_data

docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  -v n8n_data:/home/node/.n8n \
  docker.n8n.io/n8nio/n8n

Access the interface at http://localhost:5678. Data persists in the named Docker volume.

Adding the Firecrawl Node ​

  1. Click the + button on the canvas
  2. Search for Firecrawl in the node panel
  3. Click Install and wait for completion
  4. Drag the node onto your workflow canvas

Credential Setup ​

  1. Select the Firecrawl node and click Create New Credential
  2. Enter a credential name (e.g., Firecrawl Production)
  3. Paste your API key from the Firecrawl dashboard
  4. Click Save -- this credential is reusable across all Firecrawl nodes

Connection Test ​

  • Set the operation to Scrape a URL and get its content
  • Enter a test URL (e.g., https://firecrawl.dev)
  • Click Test step to verify connectivity

Available Operations ​

OperationFunctionBest For
Scrape URLExtract single-page content in markdown/HTML/summaryArticles, product monitoring, blog extraction
SearchWeb/news/image search with optional content scrapingResearch automation, trend discovery, lead generation
Crawl WebsiteRecursive multi-page scraping with URL filteringDocumentation extraction, site archiving
Map WebsiteDiscover all URLs without extracting contentSitemap generation, link discovery
Extract DataAI-powered structured extraction via custom promptsDatabase building, structured data gathering
Batch ScrapeParallel processing of multiple URLsBulk collection, large-scale projects
Agent (Sync)AI autonomously browses and extracts per promptComplex multi-page research
Agent (Async)Returns a job ID; use Get Agent Status for resultsLong-running extraction tasks

Trigger Types ​

TriggerDescription
Schedule TriggerTime-based execution via cron scheduling (hourly, daily, weekly)
Manual TriggerOn-demand execution from the n8n UI
WebhookEvent-triggered execution from external systems

Example Workflows ​

Workflow 1: Daily Blog Summary to Telegram ​

Pipeline: Schedule Trigger --> Firecrawl (Scrape + Summary) --> Telegram

Configuration:

  • Trigger: Daily at 9:00 AM
  • Firecrawl URL: https://www.firecrawl.dev/blog/category/product-updates
  • Format: Select Summary
  • Telegram: Map the summary field to the message text

Workflow 2: AI News Search with Formatting ​

Pipeline: Manual Trigger --> Firecrawl (Search) --> Code Node --> Telegram

The Code node formats results into a readable message:

javascript
const results = $input.all();
let message = "Latest AI News:\n\n";

results.forEach((item) => {
  item.json.data.web.forEach((article, index) => {
    message += `${index + 1}. ${article.title}\n`;
    message += `${article.description}\n`;
    message += `${article.url}\n\n`;
  });
});

return [{ json: { message } }];

Workflow 3: AI-Powered News Analysis ​

Pipeline: Manual Trigger --> Firecrawl (Search) --> Code --> AI Agent (OpenAI) --> Telegram

This adds an LLM analysis step between search and delivery:

  1. Firecrawl searches for news on your topic
  2. A Code node formats the raw results
  3. An AI Agent node analyzes and summarizes the content
  4. The summary is sent to Telegram

AI Agent configuration:

  • LLM Provider: OpenAI
  • Model: gpt-4-mini (cost-effective) or gpt-4 (advanced)
  • System Prompt: You are an AI news analyst. Analyze provided articles and create concise summaries highlighting important developments, trends, and implications.

Common Workflow Patterns ​

PatternStructure
Schedule-based monitoringSchedule Trigger --> Firecrawl --> Notification
Manual researchManual Trigger --> Firecrawl --> AI Analysis --> Export
Real-time alertsWebhook --> Firecrawl --> Conditional Branch --> Action
Batch processingManual/Schedule --> Batch Scrape --> Data Transform --> Database

Error Handling ​

  • Add Error Trigger nodes to capture failures and send notifications
  • Enable Continue On Fail for non-critical nodes in the chain
  • Review the Executions tab regularly for debugging
  • Verify data at each step by clicking individual node outputs

Performance Tips ​

  • Use Batch Scrape instead of loops when processing multiple URLs
  • Respect target site rate limits to avoid blocks
  • Schedule intensive operations during off-peak hours
  • Cache data when possible to reduce API credit consumption

Importing Community Templates ​

  1. Browse the n8n community template library for Firecrawl workflows
  2. Click the Import button for your n8n instance
  3. Configure credentials for each Firecrawl and output node
  4. Customize parameters for your specific use case
  5. Activate the workflow

TIP

Test workflows manually using the Execute button before enabling scheduled triggers. This lets you inspect output at every node.

Resources ​