Loading...

Please wait while we prepare your experience

AI AUTOMATION MASTERY

N8N AI Integration: OpenAI GPT & ChatGPT Complete Workflows

Automate content creation, customer support, and data analysis with AI. $12K/month cost savings with intelligent automation.

$12K/mo

Cost Savings

70%

Ticket Reduction

100+

Posts/Month

24/7

AI Support

Real Case Study: Digital Marketing Agency

Before N8N + AI Automation:

• $12,000/month on freelance writers
• 50+ hours/week on content creation
• 24-48 hour customer support response time
• Manual data analysis (15 hours/week)
• Inconsistent blog post quality
• Limited content output (20 posts/month)

After N8N + GPT Automation:

• $200/month on OpenAI API costs
• 10 hours/week on content review/editing
• Instant AI-powered support responses
• Automated data insights (real-time)
• Consistent, brand-aligned content
• 100+ posts/month with AI assistance

Result: $11,800/month saved + 5x content output increase

💰 API Cost Warning: Token Usage Optimization

OpenAI charges by token usage. Here's how to minimize costs:

• GPT-4: $0.03/1K prompt tokens, $0.06/1K completion tokens (expensive, use sparingly)
• GPT-3.5-Turbo: $0.0005/1K tokens (99% cheaper, use for most tasks)
• Set max_tokens: Limit response length to prevent runaway costs
• Cache responses: Store common queries in database to avoid re-processing
• Use system prompts: Shorter prompts = lower costs
• Batch requests: Process multiple items in one API call when possible

Cost Example: 100 blog posts/month

GPT-4: ~2000 tokens/post × 100 posts × $0.09/1K = $180/month

GPT-3.5-Turbo: ~2000 tokens/post × 100 posts × $0.001/1K = $2/month

Workflow 1: AI Blog Post Generation

✅ Use Case: Generate SEO-Optimized Blog Posts

Automatically generate blog post drafts from keywords, publish to WordPress/Ghost, and track performance.

Input Keywords (Airtable/Google Sheets)

Store blog topics in a spreadsheet.

Node: Airtable Trigger

Table: Blog Topics

Trigger: New Record Created

Fields: keyword, target_audience, tone

Generate Blog Outline (GPT-3.5)

Create structured outline first (cheaper, better results).

Node: OpenAI

Model: gpt-3.5-turbo

Prompt:

Create a detailed blog post outline for:

Topic: {{ $json.keyword }}

Audience: {{ $json.target_audience }}

Tone: {{ $json.tone }}

Include:

- Hook intro

- 5-7 main sections

- FAQ section

- Conclusion with CTA

Max tokens: 500

Generate Full Content (GPT-4 for Quality)

Expand outline into complete article.

Node: OpenAI

Model: gpt-4 (for best quality)

Prompt:

Write a complete blog post based on this outline:

Requirements:

- 1500-2000 words

- SEO-optimized for keyword: {{ $json.keyword }}

- Include internal links

- Conversational tone

- Add meta description

Max tokens: 3000

Generate Featured Image (DALL-E 3)

Create custom blog post image.

Node: HTTP Request (OpenAI DALL-E API)

Method: POST

URL: https://api.openai.com/v1/images/generations

Body:

{

"model": "dall-e-3",

"prompt": "Professional blog header image for: {{ $json.keyword }}",

"size": "1792x1024",

"quality": "standard"

}

Upload Image to Cloud Storage

Node: HTTP Request

Download image from DALL-E URL

↓

Node: S3/Cloudinary Upload

Upload image to your CDN

Publish to WordPress (Draft Mode)

Node: WordPress

Action: Create Post

Title: {{ $json.title }}

Content: {{ $json.content }}

Featured Image: {{ $json.image_url }}

Status: draft (for human review)

Meta Description: {{ $json.meta_description }}

Notify Editor for Review

Node: Slack

Channel: #content-review

Message:

📝 New AI-generated draft ready for review!

Topic: {{ $json.keyword }}

Word count: {{ $json.word_count }}

Review link: {{ $json.wordpress_url }}

Prompt Engineering Tips:

• Be specific: Include target word count, tone, and keywords
• Use examples: "Write like this: [example paragraph]"
• Set constraints: "No more than 2000 words"
• Request structure: "Use H2 headings for each section"
• Iterate: Generate outline first, then expand (better results)

Workflow 2: GPT-Powered Customer Support (70% Ticket Reduction)

✅ Use Case: Instant AI Support Responses

Auto-respond to customer emails with GPT, escalate complex issues to humans. 70% of tickets resolved automatically.

Trigger: New Support Email

Node: Gmail Trigger

Filter: to:support@yourcompany.com

Node: Webhook (from Zendesk/Intercom)

Classify Ticket (GPT-3.5)

Node: OpenAI

Model: gpt-3.5-turbo

Prompt:

Classify this support ticket:

Categories:

- billing_question

- technical_issue

- feature_request

- refund_request

- general_inquiry

Return only the category name.

Max tokens: 10

Fetch Relevant Knowledge Base Articles

Node: HTTP Request (Your Knowledge Base API)

Search: {{ $json.ticket_category }}

Return top 3 articles

Generate Response (GPT-4)

Node: OpenAI

Model: gpt-4 (better accuracy for support)

Prompt:

You are a helpful customer support agent.

Customer question:

Relevant knowledge base articles:

Write a helpful, friendly response.

If you cannot answer confidently, say:

"I'll escalate this to a human agent."

Max tokens: 500

Check Confidence Level (Switch Node)

Node: IF

Condition: response contains "escalate"

IF TRUE → Route to human agent

IF FALSE → Send automated response

Send Response or Escalate

Path A: Auto-respond

Node: Gmail (Send Reply)

To: {{ $json.customer_email }}

Body: {{ $json.gpt_response }}

Footer: "This is an AI-assisted response."

Path B: Escalate to human

Node: Zendesk (Create Ticket)

Priority: high

Assignee: support_team

Log Interaction (Analytics)

Node: Google Sheets

Append row:

- timestamp

- ticket_category

- auto_resolved (yes/no)

- customer_email

- response_tokens_used

Support Automation Results:

Tickets Auto-Resolved:

• Billing questions: 85%
• General inquiries: 90%
• Feature requests: 60%
• Technical issues: 40%

Average Response Time:

• Before: 4.5 hours
• After: 2 minutes
• Customer satisfaction: +35%
• Support costs: -60%

Workflow 3: Automated Image Generation with DALL-E 3

✅ Use Case: Generate Marketing Images at Scale

Automatically generate product images, social media graphics, and marketing materials with DALL-E 3.

Input Image Prompts (Airtable)

Node: Airtable Trigger

Table: Image Requests

Fields: prompt, style, size, project_name

Generate Image (DALL-E 3 API)

Node: HTTP Request
Method: POST
URL: https://api.openai.com/v1/images/generations
Headers:
  Authorization: Bearer YOUR_OPENAI_API_KEY
  Content-Type: application/json
Body:
{
  "model": "dall-e-3",
  "prompt": "{{ $json.prompt }}",
  "size": "{{ $json.size }}",
  "quality": "standard",
  "n": 1
}

Cost: $0.040 per image (1024x1024)
      $0.080 per image (1792x1024)

Download Image from URL

Node: HTTP Request

Method: GET

URL: {{ $json.data[0].url }}

Response Format: file

Upload to Cloud Storage (S3/Cloudinary)

Node: AWS S3

Operation: Upload

Bucket: your-images-bucket

File Name: {{ $json.project_name }}_{{ $now }}.png

ACL: public-read

Update Airtable with Image URL

Node: Airtable (Update Record)

Record ID: {{ $json.record_id }}

Fields:

image_url: {{ $json.s3_url }}

status: "completed"

generated_at: {{ $now }}

Notify Team (Slack)

Node: Slack

Channel: #marketing

Message:

🎨 New image generated!

Project: {{ $json.project_name }}

Prompt: {{ $json.prompt }}

View: {{ $json.s3_url }}

DALL-E 3 Cost Optimization:

• Use 1024x1024: $0.040/image (cheapest option)
• Use "standard" quality: "hd" costs 2x more
• Batch requests: Generate multiple variations at once
• Cache prompts: Save successful prompts for reuse
• Monthly budget: 100 images = $4-8/month

Complete JSON Workflow: Blog Post Generator

{
  "name": "AI Blog Post Generator",
  "nodes": [
    {
      "id": "1",
      "type": "n8n-nodes-base.airtableTrigger",
      "name": "New Blog Topic",
      "parameters": {
        "base": "appXXXXXXXXXXXXXX",
        "table": "Blog Topics",
        "triggerField": "Status",
        "triggerValue": "Ready to Generate"
      }
    },
    {
      "id": "2",
      "type": "n8n-nodes-base.openAi",
      "name": "Generate Outline",
      "parameters": {
        "model": "gpt-3.5-turbo",
        "messages": {
          "values": [
            {
              "role": "system",
              "content": "You are a professional content strategist."
            },
            {
              "role": "user",
              "content": "Create blog outline for: ={{ $json.keyword }}"
            }
          ]
        },
        "maxTokens": 500,
        "temperature": 0.7
      }
    },
    {
      "id": "3",
      "type": "n8n-nodes-base.openAi",
      "name": "Generate Full Content",
      "parameters": {
        "model": "gpt-4",
        "messages": {
          "values": [
            {
              "role": "user",
              "content": "Write complete blog post from outline: ={{ $node['Generate Outline'].json.choices[0].message.content }}"
            }
          ]
        },
        "maxTokens": 3000
      }
    },
    {
      "id": "4",
      "type": "n8n-nodes-base.httpRequest",
      "name": "Generate DALL-E Image",
      "parameters": {
        "method": "POST",
        "url": "https://api.openai.com/v1/images/generations",
        "authentication": "headerAuth",
        "jsonParameters": true,
        "bodyParametersJson": {
          "model": "dall-e-3",
          "prompt": "Professional header for: ={{ $json.keyword }}",
          "size": "1792x1024"
        }
      }
    },
    {
      "id": "5",
      "type": "n8n-nodes-base.httpRequest",
      "name": "Download Image",
      "parameters": {
        "method": "GET",
        "url": "={{ $json.data[0].url }}",
        "responseFormat": "file"
      }
    },
    {
      "id": "6",
      "type": "n8n-nodes-base.awsS3",
      "name": "Upload to S3",
      "parameters": {
        "operation": "upload",
        "bucket": "blog-images",
        "fileName": "={{ $json.keyword }}.png",
        "acl": "public-read"
      }
    },
    {
      "id": "7",
      "type": "n8n-nodes-base.wordpress",
      "name": "Publish to WordPress",
      "parameters": {
        "operation": "create",
        "resource": "post",
        "title": "={{ $json.title }}",
        "content": "={{ $node['Generate Full Content'].json.choices[0].message.content }}",
        "featuredMedia": "={{ $json.s3_url }}",
        "status": "draft"
      }
    },
    {
      "id": "8",
      "type": "n8n-nodes-base.slack",
      "name": "Notify Team",
      "parameters": {
        "channel": "#content",
        "text": "New draft ready: ={{ $json.wordpress_url }}"
      }
    }
  ],
  "connections": {
    "New Blog Topic": { "main": [[{ "node": "Generate Outline" }]] },
    "Generate Outline": { "main": [[{ "node": "Generate Full Content" }]] },
    "Generate Full Content": { "main": [[{ "node": "Generate DALL-E Image" }]] },
    "Generate DALL-E Image": { "main": [[{ "node": "Download Image" }]] },
    "Download Image": { "main": [[{ "node": "Upload to S3" }]] },
    "Upload to S3": { "main": [[{ "node": "Publish to WordPress" }]] },
    "Publish to WordPress": { "main": [[{ "node": "Notify Team" }]] }
  }
}

How to Import This Workflow:

Copy the JSON above
In N8N, click "Import from URL or File"
Paste the JSON
Update credentials (Airtable, OpenAI, S3, WordPress)
Test with a sample topic
Activate the workflow

More AI Automation Workflows

4. CSV Data Analysis

Prompt Engineering Tip

Upload CSV → GPT analyzes trends → Generates insights report → Sends to Slack

Prompt: "Analyze this sales data and identify:

- Top 5 products by revenue

- Month-over-month growth trends

- Underperforming categories

- Recommendations for improvement"

5. Email Response Drafting

Use Case

Incoming email → GPT drafts response → Human review → Send with one click

Workflow: Gmail Trigger

→ GPT generates draft

→ Save as Gmail draft

→ Slack notification: "Review draft"

6. Social Media Caption Generator

Prompt Engineering Tip

Upload image → GPT analyzes → Generates 5 caption variations → Schedule posts

Prompt: "Create 5 Instagram captions for this image:

1. Witty & humorous

2. Inspirational quote

3. Question to drive engagement

4. Product-focused CTA

5. Storytelling narrative"

7. Sentiment Analysis

Use Case

Customer feedback → GPT sentiment score → Alert if negative → Route to support lead

Prompt: "Rate this feedback (1-10):

Return only the number."

IF score < 5 → Alert team

Frequently Asked Questions

1. How much does OpenAI API cost for business automation?

It varies by usage. GPT-3.5-Turbo costs about $0.001/1K tokens (very cheap). GPT-4 is $0.03-0.06/1K tokens. For 100 blog posts/month using GPT-3.5, expect $2-5/month. For customer support (1000 tickets/month), expect $10-30/month.

2. Should I use GPT-4 or GPT-3.5-Turbo?

Use GPT-3.5-Turbo for most tasks (90%+ accuracy, 99% cheaper). Use GPT-4 only for critical tasks requiring highest accuracy: customer support, legal content, complex data analysis. Our clients use 80% GPT-3.5, 20% GPT-4.

3. How do I prevent GPT from generating incorrect information?

Use system prompts to set guidelines. Provide context from your knowledge base. Always include: "If you're not 100% confident, say 'I'll escalate this to a human.'" Have humans review critical outputs. Use GPT-4 for higher accuracy.

4. Can I integrate GPT with my custom data/knowledge base?

Yes! Two approaches: (1) Fetch relevant data from your KB and include it in the prompt context. (2) Use OpenAI's Assistants API with file uploads (more expensive but powerful). Most N8N users use approach 1.

5. How do I handle OpenAI API rate limits?

OpenAI has tiered rate limits based on usage. New accounts: 3 requests/min (GPT-3.5), 500 requests/day (GPT-4). As you use more, limits increase. Use N8N's Rate Limit node to throttle requests and Queue node for high-volume workflows.

6. Is AI-generated content safe for SEO?

Google doesn't penalize AI content if it's high-quality and helpful. Key: Always have humans review/edit AI drafts, add personal insights, ensure accuracy, and avoid duplicate content. Use AI as a first draft tool, not a final product.

7. Can GPT replace my entire support team?

No, but it can handle 60-80% of routine inquiries, allowing your team to focus on complex issues. Always have a human escalation path. Our clients reduced support staff from 5 to 2 agents, not to zero.

8. How do I track ROI on AI automation?

Track: (1) Time saved (hours/week), (2) API costs vs previous solution costs, (3) Quality metrics (customer satisfaction, error rates), (4) Output volume increase. Log all API calls in Google Sheets for cost analysis.

Related Resources

N8N Beginner's Guide

Start here if you're new to N8N

Advanced Webhook Workflows

Master API integrations with N8N

Error Handling & Monitoring

Keep your AI workflows running 24/7

Ready to Build AI-Powered Workflows?

Get our complete N8N AI Automation Template Pack with pre-built GPT workflows and prompt libraries.

Get AI Automation Templates

AI AUTOMATION MASTERY

N8N AI Integration: OpenAI GPT & ChatGPT Complete Workflows

Automate content creation, customer support, and data analysis with AI. $12K/month cost savings with intelligent automation.

$12K/mo

Cost Savings

70%

Ticket Reduction

100+

Posts/Month

24/7

AI Support

Real Case Study: Digital Marketing Agency

Before N8N + AI Automation:

• $12,000/month on freelance writers
• 50+ hours/week on content creation
• 24-48 hour customer support response time
• Manual data analysis (15 hours/week)
• Inconsistent blog post quality
• Limited content output (20 posts/month)

After N8N + GPT Automation:

• $200/month on OpenAI API costs
• 10 hours/week on content review/editing
• Instant AI-powered support responses
• Automated data insights (real-time)
• Consistent, brand-aligned content
• 100+ posts/month with AI assistance

Result: $11,800/month saved + 5x content output increase

💰 API Cost Warning: Token Usage Optimization

OpenAI charges by token usage. Here's how to minimize costs:

• GPT-4: $0.03/1K prompt tokens, $0.06/1K completion tokens (expensive, use sparingly)
• GPT-3.5-Turbo: $0.0005/1K tokens (99% cheaper, use for most tasks)
• Set max_tokens: Limit response length to prevent runaway costs
• Cache responses: Store common queries in database to avoid re-processing
• Use system prompts: Shorter prompts = lower costs
• Batch requests: Process multiple items in one API call when possible

Cost Example: 100 blog posts/month

GPT-4: ~2000 tokens/post × 100 posts × $0.09/1K = $180/month

GPT-3.5-Turbo: ~2000 tokens/post × 100 posts × $0.001/1K = $2/month

Workflow 1: AI Blog Post Generation

✅ Use Case: Generate SEO-Optimized Blog Posts

Automatically generate blog post drafts from keywords, publish to WordPress/Ghost, and track performance.

Input Keywords (Airtable/Google Sheets)

Store blog topics in a spreadsheet.

Node: Airtable Trigger

Table: Blog Topics

Trigger: New Record Created

Fields: keyword, target_audience, tone

Generate Blog Outline (GPT-3.5)

Create structured outline first (cheaper, better results).

Node: OpenAI

Model: gpt-3.5-turbo

Prompt:

Create a detailed blog post outline for:

Topic: {{ $json.keyword }}

Audience: {{ $json.target_audience }}

Tone: {{ $json.tone }}

Include:

- Hook intro

- 5-7 main sections

- FAQ section

- Conclusion with CTA

Max tokens: 500

Generate Full Content (GPT-4 for Quality)

Expand outline into complete article.

Node: OpenAI

Model: gpt-4 (for best quality)

Prompt:

Write a complete blog post based on this outline:

Requirements:

- 1500-2000 words

- SEO-optimized for keyword: {{ $json.keyword }}

- Include internal links

- Conversational tone

- Add meta description

Max tokens: 3000

Generate Featured Image (DALL-E 3)

Create custom blog post image.

Node: HTTP Request (OpenAI DALL-E API)

Method: POST

URL: https://api.openai.com/v1/images/generations

Body:

{

"model": "dall-e-3",

"prompt": "Professional blog header image for: {{ $json.keyword }}",

"size": "1792x1024",

"quality": "standard"

}

Upload Image to Cloud Storage

Node: HTTP Request

Download image from DALL-E URL

↓

Node: S3/Cloudinary Upload

Upload image to your CDN

Publish to WordPress (Draft Mode)

Node: WordPress

Action: Create Post

Title: {{ $json.title }}

Content: {{ $json.content }}

Featured Image: {{ $json.image_url }}

Status: draft (for human review)

Meta Description: {{ $json.meta_description }}

Notify Editor for Review

Node: Slack

Channel: #content-review

Message:

📝 New AI-generated draft ready for review!

Topic: {{ $json.keyword }}

Word count: {{ $json.word_count }}

Review link: {{ $json.wordpress_url }}

Prompt Engineering Tips:

• Be specific: Include target word count, tone, and keywords
• Use examples: "Write like this: [example paragraph]"
• Set constraints: "No more than 2000 words"
• Request structure: "Use H2 headings for each section"
• Iterate: Generate outline first, then expand (better results)

Workflow 2: GPT-Powered Customer Support (70% Ticket Reduction)

✅ Use Case: Instant AI Support Responses

Auto-respond to customer emails with GPT, escalate complex issues to humans. 70% of tickets resolved automatically.

Trigger: New Support Email

Node: Gmail Trigger

Filter: to:support@yourcompany.com

Node: Webhook (from Zendesk/Intercom)

Classify Ticket (GPT-3.5)

Node: OpenAI

Model: gpt-3.5-turbo

Prompt:

Classify this support ticket:

Categories:

- billing_question

- technical_issue

- feature_request

- refund_request

- general_inquiry

Return only the category name.

Max tokens: 10

Fetch Relevant Knowledge Base Articles

Node: HTTP Request (Your Knowledge Base API)

Search: {{ $json.ticket_category }}

Return top 3 articles

Generate Response (GPT-4)

Node: OpenAI

Model: gpt-4 (better accuracy for support)

Prompt:

You are a helpful customer support agent.

Customer question:

Relevant knowledge base articles:

Write a helpful, friendly response.

If you cannot answer confidently, say:

"I'll escalate this to a human agent."

Max tokens: 500

Check Confidence Level (Switch Node)

Node: IF

Condition: response contains "escalate"

IF TRUE → Route to human agent

IF FALSE → Send automated response

Send Response or Escalate

Path A: Auto-respond

Node: Gmail (Send Reply)

To: {{ $json.customer_email }}

Body: {{ $json.gpt_response }}

Footer: "This is an AI-assisted response."

Path B: Escalate to human

Node: Zendesk (Create Ticket)

Priority: high

Assignee: support_team

Log Interaction (Analytics)

Node: Google Sheets

Append row:

- timestamp

- ticket_category

- auto_resolved (yes/no)

- customer_email

- response_tokens_used

Support Automation Results:

Tickets Auto-Resolved:

• Billing questions: 85%
• General inquiries: 90%
• Feature requests: 60%
• Technical issues: 40%

Average Response Time:

• Before: 4.5 hours
• After: 2 minutes
• Customer satisfaction: +35%
• Support costs: -60%

Workflow 3: Automated Image Generation with DALL-E 3

✅ Use Case: Generate Marketing Images at Scale

Automatically generate product images, social media graphics, and marketing materials with DALL-E 3.

Input Image Prompts (Airtable)

Node: Airtable Trigger

Table: Image Requests

Fields: prompt, style, size, project_name

Generate Image (DALL-E 3 API)

Node: HTTP Request
Method: POST
URL: https://api.openai.com/v1/images/generations
Headers:
  Authorization: Bearer YOUR_OPENAI_API_KEY
  Content-Type: application/json
Body:
{
  "model": "dall-e-3",
  "prompt": "{{ $json.prompt }}",
  "size": "{{ $json.size }}",
  "quality": "standard",
  "n": 1
}

Cost: $0.040 per image (1024x1024)
      $0.080 per image (1792x1024)

Download Image from URL

Node: HTTP Request

Method: GET

URL: {{ $json.data[0].url }}

Response Format: file

Upload to Cloud Storage (S3/Cloudinary)

Node: AWS S3

Operation: Upload

Bucket: your-images-bucket

File Name: {{ $json.project_name }}_{{ $now }}.png

ACL: public-read

Update Airtable with Image URL

Node: Airtable (Update Record)

Record ID: {{ $json.record_id }}

Fields:

image_url: {{ $json.s3_url }}

status: "completed"

generated_at: {{ $now }}

Notify Team (Slack)

Node: Slack

Channel: #marketing

Message:

🎨 New image generated!

Project: {{ $json.project_name }}

Prompt: {{ $json.prompt }}

View: {{ $json.s3_url }}

DALL-E 3 Cost Optimization:

• Use 1024x1024: $0.040/image (cheapest option)
• Use "standard" quality: "hd" costs 2x more
• Batch requests: Generate multiple variations at once
• Cache prompts: Save successful prompts for reuse
• Monthly budget: 100 images = $4-8/month

Complete JSON Workflow: Blog Post Generator

{
  "name": "AI Blog Post Generator",
  "nodes": [
    {
      "id": "1",
      "type": "n8n-nodes-base.airtableTrigger",
      "name": "New Blog Topic",
      "parameters": {
        "base": "appXXXXXXXXXXXXXX",
        "table": "Blog Topics",
        "triggerField": "Status",
        "triggerValue": "Ready to Generate"
      }
    },
    {
      "id": "2",
      "type": "n8n-nodes-base.openAi",
      "name": "Generate Outline",
      "parameters": {
        "model": "gpt-3.5-turbo",
        "messages": {
          "values": [
            {
              "role": "system",
              "content": "You are a professional content strategist."
            },
            {
              "role": "user",
              "content": "Create blog outline for: ={{ $json.keyword }}"
            }
          ]
        },
        "maxTokens": 500,
        "temperature": 0.7
      }
    },
    {
      "id": "3",
      "type": "n8n-nodes-base.openAi",
      "name": "Generate Full Content",
      "parameters": {
        "model": "gpt-4",
        "messages": {
          "values": [
            {
              "role": "user",
              "content": "Write complete blog post from outline: ={{ $node['Generate Outline'].json.choices[0].message.content }}"
            }
          ]
        },
        "maxTokens": 3000
      }
    },
    {
      "id": "4",
      "type": "n8n-nodes-base.httpRequest",
      "name": "Generate DALL-E Image",
      "parameters": {
        "method": "POST",
        "url": "https://api.openai.com/v1/images/generations",
        "authentication": "headerAuth",
        "jsonParameters": true,
        "bodyParametersJson": {
          "model": "dall-e-3",
          "prompt": "Professional header for: ={{ $json.keyword }}",
          "size": "1792x1024"
        }
      }
    },
    {
      "id": "5",
      "type": "n8n-nodes-base.httpRequest",
      "name": "Download Image",
      "parameters": {
        "method": "GET",
        "url": "={{ $json.data[0].url }}",
        "responseFormat": "file"
      }
    },
    {
      "id": "6",
      "type": "n8n-nodes-base.awsS3",
      "name": "Upload to S3",
      "parameters": {
        "operation": "upload",
        "bucket": "blog-images",
        "fileName": "={{ $json.keyword }}.png",
        "acl": "public-read"
      }
    },
    {
      "id": "7",
      "type": "n8n-nodes-base.wordpress",
      "name": "Publish to WordPress",
      "parameters": {
        "operation": "create",
        "resource": "post",
        "title": "={{ $json.title }}",
        "content": "={{ $node['Generate Full Content'].json.choices[0].message.content }}",
        "featuredMedia": "={{ $json.s3_url }}",
        "status": "draft"
      }
    },
    {
      "id": "8",
      "type": "n8n-nodes-base.slack",
      "name": "Notify Team",
      "parameters": {
        "channel": "#content",
        "text": "New draft ready: ={{ $json.wordpress_url }}"
      }
    }
  ],
  "connections": {
    "New Blog Topic": { "main": [[{ "node": "Generate Outline" }]] },
    "Generate Outline": { "main": [[{ "node": "Generate Full Content" }]] },
    "Generate Full Content": { "main": [[{ "node": "Generate DALL-E Image" }]] },
    "Generate DALL-E Image": { "main": [[{ "node": "Download Image" }]] },
    "Download Image": { "main": [[{ "node": "Upload to S3" }]] },
    "Upload to S3": { "main": [[{ "node": "Publish to WordPress" }]] },
    "Publish to WordPress": { "main": [[{ "node": "Notify Team" }]] }
  }
}

How to Import This Workflow:

Copy the JSON above
In N8N, click "Import from URL or File"
Paste the JSON
Update credentials (Airtable, OpenAI, S3, WordPress)
Test with a sample topic
Activate the workflow

More AI Automation Workflows

4. CSV Data Analysis

Prompt Engineering Tip

Upload CSV → GPT analyzes trends → Generates insights report → Sends to Slack

Prompt: "Analyze this sales data and identify:

- Top 5 products by revenue

- Month-over-month growth trends

- Underperforming categories

- Recommendations for improvement"

5. Email Response Drafting

Use Case

Incoming email → GPT drafts response → Human review → Send with one click

Workflow: Gmail Trigger

→ GPT generates draft

→ Save as Gmail draft

→ Slack notification: "Review draft"

6. Social Media Caption Generator

Prompt Engineering Tip

Upload image → GPT analyzes → Generates 5 caption variations → Schedule posts

Prompt: "Create 5 Instagram captions for this image:

1. Witty & humorous

2. Inspirational quote

3. Question to drive engagement

4. Product-focused CTA

5. Storytelling narrative"

7. Sentiment Analysis

Use Case

Customer feedback → GPT sentiment score → Alert if negative → Route to support lead

Prompt: "Rate this feedback (1-10):

Return only the number."

IF score < 5 → Alert team

Frequently Asked Questions

1. How much does OpenAI API cost for business automation?

2. Should I use GPT-4 or GPT-3.5-Turbo?

3. How do I prevent GPT from generating incorrect information?

4. Can I integrate GPT with my custom data/knowledge base?

5. How do I handle OpenAI API rate limits?

6. Is AI-generated content safe for SEO?

7. Can GPT replace my entire support team?

8. How do I track ROI on AI automation?

Ready to Build AI-Powered Workflows?

Get our complete N8N AI Automation Template Pack with pre-built GPT workflows and prompt libraries.

Get AI Automation Templates