Workflow Templates

What it does Captures token usage and cost from your AI Agent/LLM. Logs model, tokens, cost, tool use, and conversation I/O to Google Sheets for simple observability and billing. Perfect for Developers adding usage monitoring to AI agents. Teams needing cost transparency in prototypes. How it works Chat Trigger collects user input for the AI Agent. A Set node injects metadata like workflow, execution, and client IDs. LangChain Code node returns a configured Chat model with a callback that reads usage metadata. The callback computes input, output, and total costs based on per‑million token prices you define. It appends token metrics to a Google Sheet via the Google Sheets Tool. The Agent records intermediate tool calls. An If node checks whether a tool was used. When tools are used, the workflow logs input, output, tool name, and metadata to an Observability sheet. How to use SELF-HOSTED N8N ONLY - the Langchain Code node is only available in the self-hosted version of n8n. It is not available in n8n cloud. Requirements Self-hosted version of n8n If you have any questions in running the workflow, see the attached video: https://youtu.be/JSulRS128MA

This workflow contains community nodes that are only compatible with the self-hosted version of n8n. What this workflow does Monitors Google Drive for new driver license image uploads Downloads and processes images using VLM Run AI OCR Extracts key information including license number, name, DOB, and dates Saves structured data to Google Sheets for instant access Setup Prerequisites: Google Drive account, VLM Run API credentials, Google Sheets access, self-hosted n8n. You need to install VLM Run community node Quick Setup: Configure Google Drive OAuth2 and create license upload folder Add VLM Run API credentials Set up Google Sheets integration for data storage Update folder/sheet IDs in workflow nodes Test with sample license images and activate Perfect for Customer onboarding and identity verification KYC compliance and document processing HR employee verification and record keeping Insurance claim processing and validation Any business requiring license data extraction Key Benefits Asynchronous processing** handles high-resolution images without timeouts Multi-format support** for JPG, PNG, PDF, HEIC, WebP formats Structured data output** ready for databases and integrations Eliminates manual entry** saving hours of data input time High accuracy OCR** with multi-state license support How to customize Extend by adding: Address and additional field extraction Data validation and error checking Integration with CRM or customer databases Email notifications for processing completion Audit trails and compliance reporting Duplicate detection and data deduplication This workflow transforms manual license data entry into an automated, accurate, and compliant process, making identity verification seamless and reliable for your business operations.

Building agentic AI workflows often requires multiple moving parts: memory management, document retrieval, vector similarity, and orchestration. Until now, these pieces had to be custom-wired. But with the new native n8n nodes for MongoDB Atlas, we reduce that overhead dramatically. With just a few clicks: Store and recall long-term memory from MongoDB Query vector embeddings stored in Atlas Vector Search Use these results in your LLM chains and automation logic In this example we present an ingestion and AI Agent flows that focus around Travel Planning. The different interest points that we want the agent to know about can be ingested into the vector store. The AI Agent will use the vector store tool to get relevant context about those points of interest if it needs to. Prerequisites MongoDB Atlas project and Cluster OpenAI Valid API Key for embeddings (can be other provider) Gemini API Key for the LLM (can be other provider) How it works: There are 2 main flows. One is ingesting flow: Gets a document from a webhook and use MongoDB Vector Atlas to embed the document title and description into points_of_interest collection. Embeddings are stored in a field named embedding Embeddings used are OpenAI's but it can be any type of supported embedders. Second flow is an AI Agent node with Chat Memory Stored in MongoDB Atlas and a Vector Search node as a tool: Chat Message Trigger**: Chatting with the AI Agent will trigger the conversation store in the MongoDB Chat Memory node. When data is necessary like a location search or details it will go to the "Vector Search" tool. Vector Search Tool** - uses Atlas Vector Search index created on the points_of_interest collection: // index name : "vector_index" // If you change an embedding provider make sure the numDimensions correspond to the model. { "fields": [ { "type": "vector", "path": "embedding", "numDimensions": 1536, "similarity": "cosine" } ] } Additional Resources MongoDB Atlas Vector Search n8n Atlas Vector Search docs

How it works This workflow demonstrates a fundamental pattern for securing a webhook by requiring an API key. It acts as a gatekeeper, checking for a valid key in the request header before allowing the request to proceed. Incoming Request: The Secured Webhook node receives an incoming POST request. It expects an API key to be sent in the x-api-key header. API Key Verification: The Check API Key node takes the key from the incoming request's header. It then makes an internal HTTP request to a second webhook (Get API Key) which acts as a mock database. This second webhook retrieves a list of registered API keys (from the Registered API Keys node) and filters it to find a match for the key that was provided. Conditional Response: If a match is found, the API Key Identified node routes the execution to the "success" path, returning a 200 OK response with the identified user's ID. If no match is found, it routes to the "unauthorized" path, returning a 401 Unauthorized error. This pattern separates the public-facing endpoint from the data source, which is a good security practice. Set up steps Setup time: ~2 minutes This workflow is designed to be a self-contained example. Set up Credentials: This workflow uses "Header Auth" for its internal communication. Go to Credentials and create a new Header Auth credential. You can use any name and value (e.g., Name: X-N8N-Auth, Value: my-secret-password). Select this credential in all four webhook/HTTP Request nodes. Add Your API Keys: Open the Registered API Keys node. This is your mock database. Edit the array to include the user_id and api_key pairs you want to authorize. Activate the workflow. Test it: Use the Test Secure Webhook node to send a request. Try it with a valid key from your list to see the success response. Change the x-api-key header to an invalid key to see the 401 Unauthorized error. For Production: Replace the mock database part of this workflow (the Get API Key webhook and Registered API Keys node) with a real database node like Supabase, Postgres, or Baserow to look up keys.

WordPress Post Auto-Categorization Workflow 💡 Click here to read detailed case study 📺 Click here to watch youtube tutorial 🎯 Purpose Automatically categorize WordPress blog posts using AI, saving hours of manual work. This workflow analyzes your post titles and assigns them to predefined categories using artificial intelligence. 🔄 What This Workflow Does • Connects to your WordPress site • Retrieves all uncategorized posts • Uses AI to analyze post titles • Automatically assigns appropriate category IDs • Updates posts with new categories • Processes dozens of posts in minutes ⚙️ Setup Requirements WordPress site with admin access Predefined categories in WordPress OpenAI API credentials (or your preferred AI provider) n8n with WordPress credentials 🛠️ Configuration Steps Add your WordPress categories (manually in WordPress) Note down category IDs Update the AI prompt with your category IDs Configure WordPress credentials in n8n Set up AI API connection 🔧 Customization Options • Modify AI prompts for different categorization criteria • Adjust for multiple category assignments • Add tag generation functionality • Customize for different content types • Add additional metadata updates ⚠️ Important Notes • Backup your WordPress database before running • Test with a few posts first • Review AI categorization results initially • Categories must be created manually first 🎁 Bonus Features • Can be modified for tag generation • Works with scheduled posts • Handles bulk processing • Maintains categorization consistency Perfect for content managers, bloggers, and website administrators looking to organize their WordPress content efficiently. #n8n #WordPress #ContentManagement #Automation #AI Created by rumjahn

Who's this for? If you own a website and need to analyze your Matomo analytics data so you can increse the number of frequent visitors If you need to create an SEO report on what are the common trends amongst your most frequent visitors If you want to grow your site based on suggestions from data Matomo is an analytics tool that can give you details of each individual visitor. Much more powerful than Google analytics. Watch youtube tutorial here Get my SEO A.I. agent system here Read more -> How to create an A.I. Agent to analyze Matomo analytics using n8n for free Here's the A.I. output: Keywords showing the most improvement: Openrouter N8N. Keywords needing attention: Ai Generated Reference Letter Obsidian Second Brain Suggested actions for improvement: Optimize for "best Docker Synology" despite stable ranking, an improvement to top 10 is an achievable goal. Since "2nd brain app for developer" is of interest to a developer. Consider writing a blog post on how the app addresses the specific pain points of developers. Use case Instead of hiring an SEO expert, I run this report weekly. It looks at the data for the past week and looks for visitors with more than 3 visits and recommends ideas to convert more visitors into frequent visitors. How it works The workflow gathers matomo analytics for the past 7 days. We then parse the data The data is sent to Openrouter and using a FREE LLM, it analyses the data. It stores the results in baserow How to use this Input your Matomo analytics credentials Input your Matomo site ID Input your Openrouter.ai credentials Input your baserow credentials You will need to create a baserow database with columns: Dates, Notes, Blog. Created by Rumjahn

Who is it for? This workflow is for marketers, sales teams, and local businesses who want to quickly collect leads (business name, phone, website, and email) from Google Maps and store them in Airtable. You can use it for real estate agents, restaurants, therapists, or any local niche. How it works Scrape Google Maps with Apify Google Maps Extractor. Clean and structure the data (name, address, phone, website). Visit each website and retrieve the raw HTML. Use GPT to extract the most relevant email from the site content. Save everything to Airtable for easy filtering and future outreach. It works for any location or keyword – just adapt the input in Apify. Requirements Before running this workflow, you’ll need: ✅ Apify account (to use the Google Maps Extractor) ✅ OpenAI API key (for GPT email extraction) ✅ Airtable account & base with the following fields: Business Name Address Website Phone Number Email Google Maps URL Airtable Structure Your Airtable base should contain these columns: Airtable Structure | Title | Street | Website | Phone Number | Email | URL | |-------------------------|-------------------------|--------------------|-----------------|------------------------|----------------------| | Paris Real Estate Agency| 10 Rue de Rivoli, Paris | https://agency.fr | +33 1 23 45 67 | contact@agency.fr | maps.google.com/... | | Example Business 2 | 25 Avenue de l’Opéra | https://example.fr | +33 1 98 76 54 | info@example.fr | maps.google.com/... | | Example Business 3 | 8 Boulevard Haussmann | https://demo.fr | +33 1 11 22 33 | contact@demo.fr | maps.google.com/... | Error Handling Missing websites:** If a business has no website, the flow skips the scraping step. No email found:** GPT returns Null if no email is detected. API rate limits:** Add a Wait node between requests to avoid Apify/OpenAI throttling. Now let’s take a detailed look at how to set up this automation, using real estate agencies in Paris as an example. Step 1 – Launch the Google Maps Scraper Start with a When clicking Execute workflow trigger to launch the flow manually. Then, add an HTTP Request node with the method set to POST. 👉 Head over to Apify: Google Maps Extractor On the page: https://apify.com/compass/google-maps-extractor Enter your business keyword (e.g., real estate agency, hairdresser, restaurant) Set the location you want to target (e.g., Paris, France) Choose how many results to fetch (e.g., 50) Optionally, use filters (only places with a website, by category, etc.) ⚠️ No matter your industry, this works — just adapt the keyword and location. Once everything is filled in: Click Run to test. Then, go to the top right → click on API. Select the API endpoints tab. Choose Run Actor synchronously and get dataset items. Copy the URL and paste it into your HTTP Request (in the URL field). Then enable: ✅ Body Content Type → JSON ✅ Specify Body Using JSON` Go back to Apify, click on the JSON tab, copy the entire code, and paste it into the JSON body field of your HTTP Request. At this point, if you run your workflow, you should see a structured output similar to this: title subTitle price categoryName address neighborhood street city postalCode ........ Step 2 – Clean and structure the data Once the raw data is fetched from Apify, we clean it up using the Edit Fields node. In this step, we manually select and rename the fields we want to keep: Title → {{ $json.title }} Address → {{ $json.address }} Website → {{ $json.website }} Phone → {{ $json.phone }} URL → {{ $json.url }}* This node lets us keep only the essentials in a clean format, ready for the next steps. On the right: a clear and usable table, easy to work with. Step 3 – Loop Over Items Now that our data is clean (see step 2), we’ll go through it item by item to handle each contact individually. The Loop Over Items node does exactly that: it takes each row from the table (each contact pulled from Apify) and runs the next steps on them, one by one. 👉 Just set a Batch Size of 20 (or more, depending on your needs). Nothing tricky here, but this step is essential to keep the flow dynamic and scalable. Step 4 – Edit Field (again) After looping through each contact one by one (thanks to Loop Over Items), we're refining the data a bit more. This time, we only want to keep the website. We use the Edit Fields node again, in Manual Mapping mode, with just: Website → {{ $json.website }} The result on the right? A clean list with only the URLs extracted from Google Maps. 🔧 This simple step helps isolate the websites so we can scrape them one by one in the next part of the flow. Step 5 – Scrape Each Website with an HTTP Request Let’s continue the flow: in the previous step, we isolated the websites into a clean list. Now, we’re going to send a request to each URL to fetch the content of the site. ➡️ To do this, we add an HTTP Request node, using the GET method, and set the URL as: {{ $json.website }} This value comes from the previous Edit Fields input This node will simply “visit” each website automatically and return the raw HTML code (as shown on the right). 📄 That’s the material we’ll use in the next step to extract email addresses (and any other useful info). We’re not reading this code manually — we’ll scan through it line by line to detect patterns that matter to us. This is a technical but crucial step: it’s how we turn a URL into real, usable data. Step 6 – Extract the Email with GPT Now that we've retrieved all the raw HTML from the websites using the HTTP Request node, it's time to analyze it. 💡 Goal: detect the most relevant email address on each site (ideally the main contact or owner). 👉 To do that, we’ll use an OpenAI node (Message a Model). Here’s how to configure it: ⚙️ Key Parameters: Model: GPT-4-1-MINI (or any GPT-4+ model available) Operation: Message a Model Resource: Text Simplify Output: ON Prompt (message you provide): Look at this website content and extract only the email I can contact this business. In your output, provide only the email and nothing else. Ideally, this email should be of the business owner, so if you have 2 or more options, try for most authoritative one. If you don't find any email, output 'Null'. Exemplary output of yours: name@examplewebsite.com {{ $json.data }} Step 7 – Save the Data in Airtable Once we’ve collected everything — the business name, address, phone number, website… and most importantly the email extracted via ChatGPT — we need to store all of this somewhere clean and organized. 👉 The best place in this workflow is Airtable. 📦 Why Airtable? Because it allows you to: Easily view and sort the leads you've scraped Filter, tag, or enrich them later And most importantly… reuse them in future automations ⚙️ What we're doing here We add an Airtable → Create Record node to insert each lead into our database. Inside this node, we manually map each field with the data collected in the previous steps: | Airtable Field | Description | Value from n8n | | -------------- | ------------------------ | ------------------------------------------ | | Title | Business name | {{ $('Edit Fields').item.json.Title }} | | Street | Full address | {{ $('Edit Fields').item.json.Address }} | | Website | Website URL | {{ $('Edit Fields').item.json.Website }} | | Phone Number | Business phone number | {{ $('Edit Fields').item.json.Phone }} | | Email | Email found by ChatGPT | {{ $json.message.content }} | | URL | Google Maps listing link | {{ $('Edit Fields').item.json.URL }} | 🧠 Reminder: we’re keeping only clean, usable data — ready to be exported, analyzed, or used in cold outreach campaigns (email, CRM, enrichment, etc.). ➡️ And the best part? You can rerun this workflow automatically every week or month to keep collecting fresh leads 🔁.

This workflow is an experiment to integrate charts in AI Agents, using the new Structured Output from OpenAI and Quickchart.io. How it works Users chat with an AI Agent. Anytime the AI Agent considers a chart is needed, it calls a tool to generate a chart OpenAI generates a chart using the Quickchart definition This object is added at the end of a Quickchart.io URL (see documentation) The url is added in the conversation via the AI Agent as markdown. Set up steps Create an OpenAI API Key Create the OpenAI credentials Use the credentials for the HTTP Request node (as Predefined Credential type) Activate your workflow Start chatting For example, you can ask the AI Agent to generate a chart about the top 5 movies at the box office Start exploring the limits Shout-out Quickchart.io is an amazing open source project that provides a free API to test. Go check them out! Example of chart

This workflow contains community nodes that are only compatible with the self-hosted version of n8n. Overview This workflow automates the process of transcribing audio files and summarizing them using OpenAI models, with the final output stored neatly in Notion. Whether you're a researcher, content creator, student, or professional, this automation saves time by converting voice recordings into actionable summaries with zero manual effort. Created by: Abdullah Dilshad Contact: iamabdullahdilshad@gmail.com Who It’s For This template is ideal for: Researchers**: Transcribe and summarize interviews, lectures, or research recordings. Content Creators**: Convert podcasts or videos into transcripts and social captions/show notes. Students**: Automatically turn lectures or study group audio into summarized notes. Professionals**: Log meeting notes and summaries directly into your Notion workspace. How It Works This four-step workflow performs the following: Step 1:* *Trigger: New Audio in Google Drive** Automatically triggers when a new audio file (MP3/WAV) is uploaded to a specified Google Drive folder.The file is then downloaded for processing. Step 2: Transcribe Audio with Whisper** The audio file is sent to OpenAI’s Whisper model for high-accuracy transcription. Step 3: Summarize Transcript with GPT-4** The transcript is passed to GPT-4, which generates a clean, concise summary. Step 4: Store Summary in Notion** A new Notion page is created with the generated summary and optional metadata (file name, upload time, etc.). Setup Instructions Step 1: Google Drive Trigger** Connect your Google Drive account. Select the folder you want to monitor. This node detects new file uploads and passes the file for download. Step 2: Download File** Downloads the new audio file for transcription. Step 3: Transcribe Recording (OpenAI Whisper) Connect your OpenAI API Key. Ensure this node receives the binary audio file. It will return the transcription as plain text. Step 3: Transcribe Recording (OpenAI Whisper)** Connect your OpenAI API Key. Ensure this node receives the binary audio file. It will return the transcription as plain text. Step 4: Summarize Transcript (GPT-4 via AI Agent)** Use your OpenAI API Key. Configure a summarization prompt like: "Summarize the following transcript in a clear and concise manner:" Connect the output from Whisper into this GPT-4 prompt. Step 5: Notion Integration** Connect your Notion account. Choose or create a database to store summaries. Map the GPT output (summary) to a "Text" or "Rich Text" property. Optionally include metadata like filename, file upload date, etc.

🎥 AI Video Generator with HeyGen 🚀 Create AI-Powered Videos in n8n with HeyGen This workflow enables you to generate realistic AI videos using HeyGen, an advanced AI platform for video automation. Simply input your text, choose an AI avatar and voice, and let HeyGen generate a high-quality video for you – all within n8n! ✅ Ideal for: Content creators & marketers 🏆 Automating personalized video messages 📩 AI-powered video tutorials & training materials 🎓 🔧 How It Works 1️⃣ Provide a text script – This will be spoken in the AI-generated video. 2️⃣ Select an Avatar & Voice – Choose from a variety of AI-generated avatars and voices. 3️⃣ Run the workflow – HeyGen processes your request and generates a video. 4️⃣ Download your video – Get the direct link to your AI-powered video! ⚡ Setup Instructions 1️⃣ Get Your HeyGen API Key Sign up for a HeyGen account. Go to your account settings and retrieve your API Key. 2️⃣ Configure n8n Credentials In n8n, create new credentials and select "Custom Auth" as the authentication type. In the Name provide : X-Api-Key And in the value paste your API key from Heygen Update the 2 http node with the right credentials. 3️⃣ Select an AI Avatar & Voice Browse available avatars & voices in your HeyGen account. Copy the Avatar ID and Voice ID for your video. 4️⃣ Run the Workflow Enter your text, avatar ID, and voice ID. Execute the workflow – your video will be generated automatically! 🎯 Why Use This Workflow? ✔️ Fully Automated – No manual editing required! ✔️ Realistic AI Avatars – Choose from a variety of digital avatars. ✔️ Seamless Integration – Works directly within your n8n workflow. ✔️ Scalable & Fast – Generate multiple videos in minutes. 🔗 Start automating AI-powered video creation today with n8n & HeyGen!

Who Is This For This workflow is ideal for content creators, bloggers, marketers, and professionals seeking to automate the creation and publication of SEO-optimized articles. It's particularly beneficial for those utilizing Notion for content management and WordPress for publishing.​ What Problem Does This Workflow Solve Manually creating SEO-friendly articles is time-consuming and requires consistent effort. This workflow streamlines the entire process—from detecting updates in Notion to publishing on WordPress—by leveraging AI for content generation, thereby reducing the time and effort involved.​ What This Workflow Does Monitor Notion Updates: Detects changes in a specified Notion database.​ AI Content Generation: Utilizes an AI model to produce an SEO-optimized article based on Notion data.​ Publish to WordPress: Automatically posts the generated article to a WordPress site.​ Email Notification: Sends an email containing the article's title and URL.​ Update Notion Database: Updates the corresponding entry in the Notion database with the article details.​ Setup Guide Prerequisites WordPress account with API access.​ API key for the AI model used.​ Notion integration with the relevant database ID.​ Credentials for the email service used (e.g., Gmail).​ Community Node Requirement: This workflow utilizes the n8n-nodes-mcp community node, which is only compatible with self-hosted instances of n8n. For more information on installing and managing community nodes, refer to the n8n documentation.​ n8n Docs Steps Import the workflow into your self-hosted n8n instance.​ Install the required community node (n8n-nodes-mcp).​ Configure API credentials for WordPress, the AI service, Notion, and the email service.​ Define necessary variables, such as the notification email address and Notion database IDs.​ Activate the workflow to automate the process.​ How to Customize This Workflow AI Prompt: Adjust the prompt used for content generation to align with your preferred tone and style.​ Article Structure: Modify the structure of the generated article by tweaking settings in the content generation node.​ Notifications: Customize the content and recipients of the emails sent post-publication.​ Notion Updates: Tailor the fields updated in Notion to suit your specific requirements.

Description Turn your Telegram bot into a powerful OCR (Optical Character Recognition) tool. This workflow allows you to send any image (like a screenshot, a photo of a document, or a picture of a sign) to your bot, and it will instantly extract and send back the text from that image. Powered by Google's advanced Gemini AI, this automation is perfect for quickly digitizing notes, saving important snippets, or avoiding manual typing. How it works This workflow performs a few high-level steps: It triggers when a new image is sent to your Telegram bot. It sends the image to the Google Gemini Vision API to be analyzed. It extracts the text found in the image. It sends the extracted text back to you as a message in Telegram. Set up steps Estimated set up time: Less than 5 minutes. The setup is straightforward. You only need to configure two credentials: Telegram Bot Credentials: To connect your bot. Google Gemini API Credentials: To use the OCR feature. You can get a free API key from Google AI Studio.