web-search-image-generation

Web Search + Image Generation

Research-powered image generation: Search the web → Generate contextual images with Google Gemini

---

📋 Overview

This workflow combines web research with AI image generation. It searches the web for current information about your topic, then uses that research to generate a contextually accurate and up-to-date infographic.

Perfect for: Creating images about current events, trending technologies, or topics requiring accurate real-world information.

---

✨ Features

• 🔍 Web research integration - Uses Firecrawl to search for current information
• 🎨 Context-aware generation - Images reflect real-world data
• 📊 Automatic prompt building - Combines topic + research into detailed prompts
• ⚡ Smart error handling - Continues even if search fails
• 🏷️ Topic-based filenames - Images named after your search topic

---

🏗️ Workflow Structure

Manual Trigger
     ↓
Set Search Topic (Code Node)
     ↓
Firecrawl Web Search
     ↓
Build Prompt (Code Node)
     ↓
Generate Image (Gemini)
     ↓
Decode Image (Code Node)
     ↓
Output: Image Binary

Nodes

1. manual_trigger: Start the workflow manually
2. set_search_topic: Define topic and description for research
3. firecrawl_search: Search web for relevant content
4. build_prompt: Combine topic + research into image prompt
5. generate_image: Generate image using Gemini
6. decode_image: Convert output to usable image file

---

📦 Prerequisites

Required APIs & Credentials

• ✅ n8n instance (self-hosted or cloud)
• ✅ Google Gemini API key (Get here)
• ✅ Firecrawl API key (Get here)

n8n Requirements

• n8n version 1.0+ with LangChain nodes
• Firecrawl integration: @mendable/n8n-nodes-firecrawl
• Google Gemini node: @n8n/n8n-nodes-langchain

---

🚀 Setup Instructions

Step 1: Get Firecrawl API Key

1. Go to firecrawl.dev
2. Sign up for a free account
3. Navigate to API Keys in your dashboard
4. Copy your API key

Free Tier: Firecrawl offers a free tier with limited searches per month

Step 2: Get Google Gemini API Key

1. Go to Google AI Studio
2. Sign in with Google account
3. Navigate to API Keys
4. Create an API key
5. Copy the key

Step 3: Configure n8n Credentials

Firecrawl Credential

1. In n8n, go to Credentials → Add Credential
2. Search for Firecrawl API
3. Enter your Firecrawl API key
4. Test and save

Google Gemini Credential

1. In n8n, go to Credentials → Add Credential
2. Search for Google PaLM API or Google Gemini
3. Enter your Google API key
4. Test and save

Step 4: Import Workflow

1. Download or copy web_search_image_generation.json
2. In n8n, click Import from File
3. Select the JSON file
4. Click Import

Step 5: Update Credential References

Update Firecrawl Node

1. Click on firecrawl_search node
2. Select your Firecrawl credential
3. Save

Update Gemini Node

1. Click on generate_image node
2. Select your Google Gemini credential
3. Save

---

🎯 How to Use

Basic Usage

1. Open the workflow in n8n
2. Click on the set_search_topic node
3. Edit TOPIC and DESCRIPTION variables:

const TOPIC = 'Your search topic here';
const DESCRIPTION = `Detailed description of what you want to research and visualize`;

4. Click Execute Workflow
5. View generated image in final node output

---

📝 Example Use Cases

Example 1: Technology Trends

const TOPIC = 'The rise of AI agents in 2026';
const DESCRIPTION = `Research how autonomous AI agents are being used in 
software development, customer service, and business automation. Focus on 
real tools and platforms that launched in 2025-2026.`;

Result: Image showing current AI agent platforms with accurate logos and names from web research.

Example 2: Market Analysis

const TOPIC = 'Remote work statistics 2026';
const DESCRIPTION = `Find latest data on remote work adoption rates, 
productivity metrics, and popular tools. Show percentage breakdowns and 
growth trends from 2020 to 2026.`;

Result: Infographic with actual statistics from recent studies and reports.

Example 3: Product Comparison

const TOPIC = 'Top project management tools comparison';
const DESCRIPTION = `Compare features, pricing, and user bases of leading 
project management platforms: Asana, Monday.com, ClickUp, Notion. 
Show current market positions and key differentiators.`;

Result: Comparison chart with accurate current information about each platform.

---

🔧 How It Works

1. Web Search Phase

The firecrawl_search node:

• Combines your topic + description into a search query
• Adds keywords like "visual infographic 2026" for relevant results
• Returns up to 6 web pages with titles and descriptions

2. Prompt Building Phase

The build_prompt node:

• Extracts titles and snippets from search results
• Combines topic, description, and research context
• Creates a detailed image generation prompt including:
  • Visual requirements (size, style, colors)
  • Research context for accuracy
  • Design specifications

3. Image Generation Phase

The generate_image node:

• Sends the enriched prompt to Google Gemini
• Receives image data (base64 or binary)
• Handles different response formats automatically

4. Decoding Phase

The decode_image node:

• Converts Gemini output to downloadable image file
• Creates filename from topic + timestamp
• Adds metadata (size, mime type, generation time)

---

🎨 Customization

Adjust Search Results Count

In firecrawl_search node, change the limit parameter:

"limit": 6  // Default (balance speed vs context)
"limit": 3  // Faster, less context
"limit": 10 // Slower, more context

Modify Prompt Template

In build_prompt node, customize the image prompt template:

const imagePrompt = `Create a professional, editorial-quality infographic...
// Add your custom instructions here
// Example: Emphasize specific colors, layouts, or styles
`;

Search Query Optimization

In firecrawl_search node, modify the query construction:

"query": "={{ $json.topic + ' ' + $json.description + ' visual infographic 2026' }}"
// Change to:
"query": "={{ $json.topic + ' statistics data visualization' }}"
// or
"query": "={{ $json.topic + ' architecture diagram blueprint' }}"

---

📊 Output

Image Files

• Format: JPEG or PNG (automatic detection)
• Filename: {topic-slug}_{timestamp}.jpg
  • Example: remote_work_statistics_2026_2026-05-05-143022.jpg
• Dimensions: As specified in prompt (default 1440x756px)

Metadata

{
  "filename": "ai_trends_2026_2026-05-05-143000.jpg",
  "mimeType": "image/jpeg",
  "size_bytes": 245890,
  "generated_at": "2026-05-05T14:30:00.000Z"
}

---

🐛 Troubleshooting

Issue: Firecrawl search fails

Symptoms: "No web research available" in logs

Solutions:

• Check Firecrawl API key is valid
• Verify you haven't exceeded API quota
• Try a more specific search topic
• Check Firecrawl service status

Note: Workflow continues even if search fails, using only topic/description

Issue: Generated image doesn't match research

Cause: Prompt building might not emphasize research context enough

Solutions:

• Edit build_prompt node to give more weight to search results
• Increase search result limit from 6 to 10
• Make topic and description more specific

Issue: Image includes outdated information

Cause: Gemini's training data vs web search results

Solutions:

• Add "2026" or "latest" to your topic
• Include "current data" in description
• Modify prompt to emphasize "use research context for accuracy"

Issue: Search returns irrelevant results

Cause: Topic too broad or vague

Solutions:

• Make topic more specific
• Add clarifying keywords to description
• Adjust the search query in firecrawl_search node

---

💡 Tips for Better Results

Search Topic Tips

✅ Good topics:

• "State of AI infrastructure in 2026"
• "Docker vs Kubernetes adoption rates 2025-2026"
• "Top 5 CRM platforms market share 2026"

❌ Avoid vague topics:

• "Technology stuff"
• "Business things"
• "Some data"

Description Tips

1. Be specific: Mention exact aspects you want researched
2. Include timeframes: "2025-2026", "latest", "current"
3. Name entities: "OpenAI GPT-4", "Microsoft Azure", "GitHub Copilot"
4. Request data types: "statistics", "comparisons", "architecture", "workflow"

Image Style Preferences

Add style instructions to your description:

• "Use modern isometric 3D style"
• "Create a professional comparison chart"
• "Design a timeline visualization"
• "Show as a hub-and-spoke diagram"

---

📈 Performance

Execution Time

• Web search: 2-5 seconds
• Prompt building: < 1 second
• Image generation: 5-15 seconds
• Total: ~10-25 seconds per execution

API Costs

• Firecrawl: Free tier available, paid plans for high volume
• Google Gemini: Usage-based pricing, check current rates at ai.google.dev/pricing

---

🔄 Next Steps

Upgrade to Enhanced Version

Try the Enhanced Prompt Engineering workflow which adds:

• AI-powered prompt optimization
• Style selection based on content structure
• Multi-variation generation
• Advanced visual analysis

Batch Processing

Modify the workflow to:

1. Read topics from a CSV or Google Sheets
2. Loop through each topic
3. Generate multiple images automatically

Integration Ideas

• Webhook trigger: Generate images from external requests
• Schedule trigger: Daily trend visualization
• Slack integration: Post generated images to channels
• Google Drive: Auto-save images to Drive folders

---

🎓 Learning Resources

Firecrawl Documentation

• Firecrawl API Docs
• Search API Reference

Google Gemini

• Gemini API Documentation
• Image Generation Guide

n8n Resources

• n8n Workflow Examples
• Working with Binary Data
• Error Handling

---

🤝 Contributing

Improvements and variations welcome! Share your customizations with the community.

---

📄 License

This workflow is provided as-is for community use. Modify as needed for your use case.

---

Part of the n8n-selfhosting-workflows collection

← Basic Version | View Enhanced Version →

Back to Workflows