Google Gemini Chat Model and HTTP Request integration

Important Notes: Check Legal Regulations: This workflow involves scraping, so ensure you comply with the legal regulations in your country before getting started. Better safe than sorry! Workflow Description: 😮‍💨 Tired of struggling with XPath, CSS selectors, or DOM specificity when scraping ? This AI-powered solution is here to simplify your workflow! With a vision-based AI Agent, you can extract data effortlessly without worrying about how the DOM is structured. This workflow leverages a vision-based AI Agent, integrated with Google Sheets, ScrapingBee, and the Gemini-1.5-Pro model, to extract structured data from webpages. The AI Agent primarily uses screenshots for data extraction but switches to HTML scraping when necessary, ensuring high accuracy. Key Features: Google Sheets Integration**: Manage URLs to scrape and store structured results. ScrapingBee**: Capture full-page screenshots and retrieve HTML data for fallback extraction. AI-Powered Data Parsing**: Use Gemini-1.5-Pro for vision-based scraping and a Structured Output Parser to format extracted data into JSON. Token Efficiency**: HTML is converted to Markdown to optimize processing costs. This template is designed for e-commerce scraping but can be customized for various use cases.

This n8n template demonstrates the beginnings of building your own n8n-powered WhatsApp chatbot! Under the hood, utilise n8n's powerful AI features to handle different message types and use an AI agent to respond to the user. A powerful tool for any use-case! How it works Incoming WhatsApp Trigger provides a way to get messages into the workflow. The message received is extracted and sent through 1 of 4 branches for processing. Each processing branch uses AI to analyse, summarize or transcribe the message so that the AI agent can understand it. The supported types are text, image, audio (voice notes) and video. The AI Agent is used to generate a response generally and uses a wikipedia tool for more complex queries. Finally, the response message is sent back to the WhatsApp user using the WhatsApp node. How to use Once you have setup and configured your WhatsApp account, you'll need to activate your workflow to start processing messages. Good to know: Large media files may negatively impact workflow performance. Requirements WhatsApp Buisness account Google Gemini for LLM. Gemini is used specifically because it can accept audio and video files whereas at time of writing, many other providers like OpenAI's GPT, do not. Customising this workflow For performance reasons, consider detecting large audio and video before sending to the LLM. Pre-processing such files may allow your agent to perform better. Go beyond and create rich and engagement customer experiences by responding using images, audio and video instead of just text!

Learning something new? Endlessly searching to find the best resources? This workflow finds top community-recommended learning resources on any topic from Hacker News, delivered to your inbox. How it works User submits a topic they want to learn via a simple form. The workflow searches for relevant "Ask HN" posts on Hacker News and extracts top-level comments. An LLM analyzes the comments and identifies the best learning resources. A personalized email is sent to the user with a Markdown formatted list of top recommendations, categorized by resource type (e.g., book, course, article) and difficulty level. Set up steps Add your Google Gemini API credentials. You'll need to create a project and enable the Generative Language API. Add your SMTP credentials for sending emails. Customize the Form and email subject (optional) Activate the workflow Screenshots for Workflow, Form and Email Built on Day-03 as part of the #100DaysOfAgenticAi Fork it, tweak it, have fun!

Workflow Description: Product Data Extractor This workflow automates the extraction of product data from Product Hunt by combining webhook interactions, HTML processing, AI-based data analysis, and structured output formatting. It is designed to handle incoming requests dynamically and return detailed JSON responses for further usage. Overview The workflow processes a product name submitted through a webhook. It fetches the corresponding Product Hunt page, extracts and analyzes inline scripts, and structures the data into a well-defined JSON format using AI tools. The final JSON response is returned to the client through the webhook. Workflow Steps Webhook Listener Node:** Receive Product Request Function:** Captures incoming requests containing the product name to process. Details:** Accepts HTTP requests and extracts the product parameter from the query string, such as <custom_webhook_url>/?product=epigram. Fetch Product HTML Node:** Fetch Product HTML Function:** Sends an HTTP request to retrieve the HTML content of the specified Product Hunt page. Details:** Constructs a dynamic URL using the product name and fetches the page data. Extract Inline Scripts Node:** Extract Inline Scripts Function:** Parses the HTML content to extract inline scripts located within the <head> section. Details:** Excludes scripts containing src attributes and validates the presence of inline scripts. Process Data with LLM Node:** Process Script with LLM Function:** Analyzes the extracted scripts using a language model to identify key product data. Details:** Processes the script to derive structured and meaningful insights. Refine Data with Google Gemini Node:** Analyze Script with Google Gemini Function:** Leverages Google Gemini AI for enhanced analysis of script data. Details:** Ensures the extracted data is precise and enriched. Format Product Data to JSON Node:** Format Product Data to JSON Function:** Structures the processed data into a clean JSON format. Details:** Defines a schema to ensure all relevant fields are included in the output. Send JSON Response to Client Node:** Send JSON Response to Client Function:** Returns the final structured JSON response to the client. Details:** Sends the response back via the same webhook that initiated the request. For example, <custom_webhook_url>. Key Features Versatile Use Cases:** This workflow can be used to gather Product Hunt data for creating blog posts or as a tool for AI agents to research products efficiently. Dynamic Processing:** Adapts to various product names through dynamic URL construction. AI Integration:** Utilizes the Gemini 1.5 8B AI model, offering reduced latency and minimal or no cost depending on the use case. Selector Independence:** Functions even if Product Hunt's DOM structure changes, as it does not rely on direct DOM selectors. Reliable Data Output:** A low temperature setting (0) and a precisely defined JSON schema ensure accurate and real data extraction. Dynamic Processing:** Adapts to various product names through dynamic URL construction. AI Integration:** Utilizes advanced language models for data extraction and refinement. Structured Output:** Ensures the output JSON adheres to a predefined schema for consistency. Error Handling:** Includes validations to handle missing or malformed data gracefully. Customization Options Limitations Dependency on Product Hunt:** Significant changes to the way Product Hunt loads data on its pages might require modifications to the workflow. Adaptability:** Even if changes occur, the workflow can be updated to maintain functionality due to its reliance on AI and not direct DOM selectors. Modify the webhook path to suit your application. Adjust the prompt for the language model to include additional fields. Extend the JSON schema to capture more data fields as needed. Expected Output Performance Metrics Response Time:** Typically ~6 seconds per product. Accuracy:** Data extracted with >95% precision due to the pre-defined JSON schema. A JSON object containing detailed information about the specified product. Below is an example of a complete response for the product Epigram: { "id": "861675", "slug": "epigram", "followersCount": 181, "name": "Epigram", "tagline": "Open-Source, Free, and AI-Powered News in Short", "reviewsRating": 0, "logoUuid": "735c2528-554c-467c-9dcf-745ee4b8bbdd.png", "postsCount": 1, "websiteUrl": "https://epigram.news", "websiteDomain": "epigram.news", "metaTitle": "Epigram - Open-source, free, and ai-powered news in short", "postName": "Epigram", "postTagline": "Open-source, free, and ai-powered news in short", "dailyRank": "3", "description": "An open-source, AI-powered news app for busy people. Stay updated with bite-sized news, real-time updates, and in-depth analysis. Experience balanced, trustworthy reporting tailored for fast-paced lifestyles in a sleek, user-friendly interface.", "pricingType": "free", "userName": "Fazle Rahman", "userHeadline": "Co-founder & CEO, Hashnode", "userUsername": "fazlerocks", "userAvatarUrl": "https://ph-avatars.imgix.net/129147/f84e1796-548b-4d6f-9dcf-745ee4b8bbdd.jpeg", "makerName1": "Fazle Rahman", "makerHeadline1": "Co-founder & CEO, Hashnode", "makerUsername1": "fazlerocks", "makerAvatarUrl1": "https://ph-avatars.imgix.net/129147/f84e1796-548b-4d6f-9dcf-745ee4b8bbdd.jpeg", "makerName2": "Sandeep Panda", "makerHeadline2": "Co-Founder @ Hashnode", "makerUsername2": "sandeepg33k", "makerAvatarUrl2": "https://ph-avatars.imgix.net/101872/80b0b618-a540-4110-a6d1-74df39675ad0.jpeg", "primaryLinkUrl": "https://epigram.news/", "media1OriginalHeight": 1080, "media1OriginalWidth": 1440, "media1ImageUuid": "ac426fd1-3854-4734-b43d-34a5e06347ea.gif", "media1MediaType": "video", "media1MetadataUrl": "https://www.loom.com/share/b1a48a9b3cac4ba89ce772a3fbcc2847?sid=75efc771-25fa-4ac0-bb1b-5e38fc447deb", "media1VideoId": "b1a48a9b3cac4ba89ce772a3fbcc2847", "media2OriginalHeight": 630, "media2OriginalWidth": 1200, "media2ImageUuid": "8521a6bd-7640-487b-abd6-29b9f65fee32", "media2MediaType": "image", "media2MetadataUrl": null, "launchState": "featured", "thumbnailImageUuid": "735c2528-554c-467c-9dcf-745ee4b8bbdd.png", "link1StoreName": "Website", "link1WebsiteName": "epigram.news", "link2StoreName": "Github", "link2WebsiteName": "github.com", "latestScore": 233, "launchDayScore": 233, "userId": "129147", "topic1": "News", "topic2": "Open Source", "topic3": "Artificial Intelligence", "weeklyRank": "24", "commentsCount": 20, "postUrl": "https://www.producthunt.com/posts/epigram" } Target Audience This workflow is ideal for developers, marketers, and data analysts seeking to automate the extraction and structuring of product data from Product Hunt for analytics, reporting, or integration with other tools.

Who is this for? This workflow is designed for businesses or developers looking to integrate voice-based chat applications with dynamic responses and conversational memory. What problem does this solve? It automates AI-powered voice conversations, maintaining context between sessions and converting speech-to-text and text-to-speech. What this workflow does: The workflow receives audio input, transcribes it using OpenAI, and processes the conversation using Google Gemini Chat Model (you can use OpenAI Chat Model). Responses are converted back to speech using ElevenLabs. Prerequisites: You'll need API keys for: OpenAI (you can obtain it from OpenAI website) ElevenLabs (you can obtain it from their website) Google Gemini (You can obtain it from Google AI Studio) Setup: Configure you API keys Ensure that the value (voice_message) in the "Path" parameter in the Webhook node is used as the name of the parameter that will contain the voice message you are sending via the HTTP Post request.

📈 Daily Financial News - Description This workflow automates the process of collecting, organizing, and delivering a daily summary of financial news by following these key steps: Scheduled Activation The workflow starts at 7:00 AM each day, triggered by the Schedule Trigger node. News Retrieval The HTTP Request node fetches the latest financial news from FT.com. Content Extraction The Extract Specific Content node scrapes targeted sections of the HTML (headlines, editor's picks, top stories, etc.) using CSS selectors to locate and capture relevant content. News Aggregation The Set Node gathers and organizes the extracted news data, preparing it for summarization. Categories like "Headline #1," "Editor's Picks," and "Europe News" are all structured into a single data block. Summarization An AI Agent (Google Gemini) takes the aggregated news data and creates a concise, HTML-formatted summary tailored to give investors an insightful market snapshot. Email Delivery Finally, the Microsoft Outlook node sends the summary via email to the designated recipient with the subject "Financial news from today." This process ensures that financial news is efficiently curated, summarized, and delivered without manual intervention.

This n8n workflow demonstrates an approach to parsing bank statement PDFs with multimodal LLMs as an alternative to traditional OCR. This allows for much more accurate data extraction from the document especially when it comes to tables and complex layouts. Multimodal Parsing is better than traditiona OCR because: It reduces complexity and overhead by avoiding the need to preprocess the document into text format such as markdown before passing to the LLM. It handles non-standard PDF formats which may produce garbled output via traditional OCR text conversion. It's orders of magnitude cheaper than premium OCR models that still require post-processing cleanup and formatting. LLMs can format to any schema or language you desire! How it works You can use the example bank statement created specifically for this workflow here: https://drive.google.com/file/d/1wS9U7MQDthj57CvEcqG_Llkr-ek6RqGA/view?usp=sharing A PDF bank statement is imported via Google Drive. For this demo, I've created a mock bank statement which includes complex table layouts of 5 columns. Typically, OCR will be unable to align the columns correctly and mistake some deposits for withdrawals. Because multimodal LLMs do not accept PDFs directly, well have to convert the PDF to a series of images. We can achieve this by using a tool such as Stirling PDF. Stirling PDF is self-hostable which is handy for sensitive data such as bank statements. Stirling PDF will return our PDF as a series of JPGs (one for each page) in a zipped file. We can use n8n's decompress node to extract the images and ensure they are ordered by using the Sort node. Next, we'll resize each page using the Edit Image node to ensure the right balance between resolution limits and processing speed. Each resized page image is then passed into the Basic LLM node which will use our multimodal LLM of choice - Gemini 1.5 Pro. In the LLM node's options, we'll add a "user message" of type binary (data) which is how we add our image data as an input. Our prompt will instruct the multimodal LLM to transcribe each page to markdown. Note, you do not need to do this - you can just ask for data points to extract directly! Our goal for this template is to demonstrate the LLMs ability to accurately read the page. Finally, with our markdown version of all pages, we can pass this to another LLM node to extract required data such as deposit line items. Requirements Google Gemini API for Multimodal LLM. Google Drive access for document storage. Stirling PDF instance for PDF to Image conversion Customising the workflow At time of writing, Gemini 1.5 Pro is the most accurate in text document parsing with a relatively low cost. If you are not using Google Gemini however you can switch to other multimodal LLMs such as OpenAI GPT or Antrophic Claude. If you don't need the markdown, simply asking what to extract directly in the LLM's prompt is also acceptable and would save a few extra steps. Not parsing any bank statements any time soon? This template also works for Invoices, inventory lists, contracts, legal documents etc.

This n8n workflow demonstrates how we can use Multimodal LLMs to parse and extract from PDF documents in n8n. In this particular scenario, we're passing a candidate's CV/resume to an AI which filters out unqualified applications. However, this sneaky candidate has added in hidden prompt to bypass our bot! Whatever will we do? No fret, using AI Vision is one approach to solve this problem... read on! How it works Our candidate's CV/Resume is a PDF downloaded via Google Drive for this demonstration. The PDF is then converted into an image PNG using a tool called Stirling PDF. Since the hidden prompt has a white font color, it is is invisible in the converted image. The image is then forwarded to a Basic LLM node to process using our multimodal model - in this example, we'll use Google's Gemini 1.5 Pro. In the Basic LLM node, we'll need to set a User Message with the type of Binary. This allows us to directly send the image file in our request. The LLM is now immune to the hidden prompt and its response is has expected. The example CV/Resume with hidden prompt can be found here: https://drive.google.com/file/d/1MORAdeev6cMcTJBV2EYALAwll8gCDRav/view?usp=sharing Requirements Google Gemini API Key. Alternatively, GPT4 will also work for this use-case. Stirling PDF or another service which can convert PDFs into images. Note for data privacy, this example uses a public API and it is recommended that you self-host and use a private instance of Stirling PDF instead. Customising the workflow Swap out the manual trigger for another trigger such as a webhook to integrate into your existing services. This example demonstrates a validation use-case ie. "does the candidate look qualified?". You can try additionally extracting data points instead such as years of experiences, previous companies etc.

Workflow Description This workflow is a powerful, fully automated web query and semantic reranking system that allows users to perform precise, detailed searches, intelligently rank search results and provide high-quality, structured output. Built with AI-powered components, the workflow leverages semantic query generation, result re-ranking, and real-time reporting to deliver actionable insights. It is particularly well-suited for real-time data retrieval, market research, and any domain requiring automated yet customizable search result processing. How It Works Webhook Integration for Input: The workflow begins with a Webhook Node that captures the user's search query as input, enabling seamless integration with other systems. Step 1: Semantic Query Generation (Powered by "Semantic Search - Query Maker"): Using AI (Google Gemini), the initial query is refined and transformed into a context-aware, expert-level search query. The process ensures that the search engine retrieves the most relevant and precise results. Step 2: Web Search Execution: A free Brave Search API processes the refined query to fetch search results, ensuring speed and cost efficiency. Step 3: Semantic Re-Ranking of Results (Powered by "Semantic Search - Result Re-Ranker"): The workflow reranks the search results based on relevance to the original question, prioritizing the most relevant URLs dynamically. Results are passed through AI-powered intelligent reranking to ensure the final output reflects optimal relevance and quality. Step 4: Structured Output Generation: Results are converted into a well-structured, organized JSON format, ranking the top 10 search results with their titles, links, and descriptions. Missing ranks (if fewer than 10 results) are handled gracefully with placeholders, ensuring consistency. Step 5: Real-Time Reporting: The reranked search results are sent back to the user or integrated system via the Webhook Node in a JSON-formatted response. Reports are highly structured and ready for downstream processing or consumption. Key Features AI-Powered Query Refinement: Transforms basic queries into detailed, expert-level search terms for optimal results. Dual-Stage Semantic Search: Combines query generation and result reranking for precise, high-relevance outputs. Top 10 Result Reranking: Dynamically ranks and organizes the top 10 results based on semantic relevance to the query. Customizable Integration: Fully modifiable for alternative APIs or integrations, such as other search engines or custom ranking logic. JSON-Formatted Structured Results: Outputs reranked results in a standardized format, ideal for integration into systems requiring machine-readable data. Webhook-Based Flexibility: Works seamlessly with Webhook inputs for easy deployment in diverse workflows. Cost-Effective API Usage: Pre-integrated with the free Brave Search API, minimizing operational costs while delivering accurate search results. Instructions for API Setup Brave Search API: Visit api.search.brave.com to obtain a free-tier API key for web search. AI Integration (Google Gemini): Visit Google AI Studio and generate an API key for semantic query generation and reranking. Webhook Configuration: Set up the input Webhook to capture search queries and the output Webhook to deliver reranked results. Why Choose This Workflow? Precision and Relevance**: Combines AI-based query generation with advanced reranking for accurate results. Fully Customizable**: Easily adapt the workflow to alternative APIs, search engines, or ranking logic. Real-Time Insights**: Provides structured, real-time output ready for immediate use. Scalable and Modular**: Ideal for businesses, researchers, and data analysts needing a robust, repeatable solution. Tags AI Workflow, Semantic Search, Query Refinement, Search Result Reranking, Real-Time Search, Web Search Automation, Google Search, Brave Search, News Search, API Integration, Market Research, Competitive Intelligence, Business Intelligence,Google Gemini, Anthropic Claude, OpenAI, GPT, LLM

This n8n workflow demonstrates how to automate image captioning tasks using Gemini 1.5 Pro - a multimodal LLM which can accept and analyse images. This is a really simple example of how easy it is to build and leverage powerful AI models in your repetitive tasks. How it works For this demo, we'll import a public image from a popular stock photography website, Pexel.com, into our workflow using the HTTP request node. With multimodal LLMs, there is little do preprocess other than ensuring the image dimensions fit within the LLMs accepted limits. Though not essential, we'll resize the image using the Edit image node to achieve fast processing. The image is used as an input to the basic LLM node by defining a "user message" entry with the binary (data) type. The LLM node has the Gemini 1.5 Pro language model attached and we'll prompt it to generate a caption title and text appropriate for the image it sees. Once generated, the generated caption text is positioning over the original image to complete the task. We can calculate the positioning relative to the amount of characters produced using the code node. An example of the combined image and caption can be found here: https://res.cloudinary.com/daglih2g8/image/upload/f_auto,q_auto/v1/n8n-workflows/l5xbb4ze4wyxwwefqmnc Requirements Google Gemini API Key. Access to Google Drive. Customising the workflow Not using Google Gemini? n8n's basic LLM node supports the standard syntax for image content for models that support it - try using GPT4o, Claude or LLava (via Ollama). Google Drive is only used for demonstration purposes. Feel free to swap this out for other triggers such as webhooks to fit your use case.

YouTube Subtitles Report Generator Overview This template enables users to generate analytical reports from YouTube video subtitles, providing insights into the thematic content of the video. Designed for efficiency and simplicity, it processes video subtitles without requiring an API key, making it an accessible solution for content analysis. The system assumes videos already have subtitles available, excluding live streams and videos without subtitle data. Key Features Trigger Webhook: Seamlessly receive video IDs through a webhook trigger. Fetch Video HTML: Retrieves the video’s HTML content directly from YouTube. Extract Subtitles URLs: Processes the HTML content to find and decode subtitle URLs. Fetch Subtitles Content: Downloads the subtitles from the decoded URLs in XML format. Generate Analytical Report: Utilizes an AI model to summarize and analyze the thematic essence of video content. The system supports models such as Google Gemini, OpenAI, and other compatible language models. The quality of the results may vary depending on the model used, with better models providing faster and more accurate summaries. The default prompt focuses on identifying and summarizing the main theme of the video while excluding content related to promotions, subscriptions, or sponsored content. Return Analytical Report: Delivers concise analytical reports as the final response to the webhook, suitable for various use cases like research, content creation, or consumption as plain text. Setup Instructions Step 1: Webhook Configuration Set up the webhook URL in your external system (e.g., app, API) to send YouTube video IDs to this workflow. Step 2: API Access Ensure that your environment has access to YouTube’s public HTML content. An API key is not required since the workflow parses publicly available data. Step 3: AI Integration Verify the connection to the AI model used for report generation. Supported models include Google Gemini and OpenAI. Note that the system can be customized by modifying the provided prompt to suit specific analysis needs. Step 4: Testing Run a sample test with a YouTube video ID to ensure subtitles are correctly retrieved and the report is generated successfully. Expected Outcomes Effortless Content Analysis:** Generate thematic reports with minimal setup. No API Key Dependency:** Simplified access by leveraging YouTube’s public HTML. Actionable Insights:** Gain valuable information on video content for business, research, or personal projects. Cost Considerations:** The execution cost depends primarily on the model used and the length of the video (affecting token usage). Leveraging the free tier of Google Gemini models is recommended to minimize costs. Main Theme Extraction:** The default prompt excludes irrelevant promotional content, providing clear and focused thematic summaries. Estimated setup time: 15–20 minutes with a ready environment. Prompt Description The workflow includes a customizable prompt used to process subtitles in XML format and generate analytical reports. The generated report includes: Title: A brief phrase capturing the essence of the main theme. Body: An analytical description of the main theme, structured into a maximum of three concise paragraphs. It focuses on summarizing key ideas, recurring themes, and connections without referencing the source explicitly (e.g., avoiding phrases like “this video”). The system also removes content related to sales, subscriptions, or promotions to maintain a clear thematic focus. Example Output Title: The Ethical Challenges of Artificial Intelligence Report: Artificial intelligence presents significant challenges in areas such as privacy, fairness, and automated decision-making. Its implementation in critical sectors such as healthcare, justice, and security has sparked debates about inherent biases and lack of transparency. Additionally, there is growing concern over the ethical implications of automation, including its impact on employment and the global economy. Finally, the importance of establishing strong regulatory frameworks is highlighted to balance technological innovation with the protection of human rights.

This n8n workflow is a proof-of-concept template exploring how we might work with multimodal LLMs and their multi-image analysis capabilities. In this demo, we compare 2 screenshots of a webpage taken at different timestamps and pass both to our multimodal LLM for a visual comparison of differences. Handling multiple binary inputs (ie. images) in an AI request is supported by n8n's basic LLM node. How it works This template is intended to run as 2 parts: first to generate the base screenshots and next to run the visual regression test which captures fresh screenshots. Starting with a list of webpages captured in a Google sheet, base screenshots are captured for each using a external web scraping service called Apify.com (I prefer Apify but feel free to use whichever web scraping service available to you) These base screenshots are uploaded to Google Drive and will be referenced later when we run our testing. Phase 2 of the workflow, we'll use a scheduled trigger to fire sometime in the future which will reuse our web scraping service to generate fresh screenshots of our desired webpages. Next, re-download our base screenshots in parallel and with both old and new captures, we'll pass these to our LLM node. In the LLM node's options, we'll define 2 "user message" inputs with the type of binary (data) for our images. Finally, we'll prompt our LLM with our testing criteria and capture the regressions detected. Note, results will vary depending on which LLM you use. A final report can be generated using the LLM's output and is uploaded to Linear. Requirements Apify.com API key for web screenshotting service Google Drive and Sheets access to store list of webpages and captures Customising this workflow Have your own preferred web screenshotting service? Feel free to swap out Apify with your service of choice. If the web screenshot is too large, it may prove difficult for the LLM to spot differences with precision. Try splitting up captures into smaller images instead.