With just a few nodes in n8n you can have a simple image editor using Google's Gemini 2.5 Flash Image model, codenamed "Nano Banana". I'm using an HTTP Request node to call the Gemini API directly, basically following the same process from https://aistudio.google.com/apikey, and with a n8n and a little base64 encoding and decoding, you've got a prompt-based image editor. Here is the JSON body example of the Gemini request for reference: `{ "contents": [ { "parts": [ { "text": "{{ $('On form submission').item.json.Prompt }}" }, { "inlineData": { "mimeType": "image/png", "data": "{{ $json.data }}" } } ] } ], "generationConfig": { "responseModalities": ["IMAGE"] } }` This was one of those immediately powerful workflows. I actually ended up using it create the thumbnail to remove the Salesforce branding from my sweatshirt after the fact. Really powerful model and really easy setup. May have to do a few more workflows using Nano Banana. Let me know if you try it out or have any other ideas you want to see! 🍌

Easily populate your Monday.com columns with data from PDFs with this mostly straightforward workflow using the Extract From File node in n8n and GPT (or your AI of choice). Relatively simple workflow that triggers from a Monday.com status change webhook, downloads the attached PDF, extracts the text, sends it to GPT to reformat the data into something that can be uploaded to Monday.com, then uses a Monday.com node to update the columns with matching data from the PDF. No more manually entering data from documents! Let me know if you have any comments or questions.

This is similar to another post where I showed how to extract data from PDFs and turn them into columns in Monday.com, but I found out that the Extract from PDF node in n8n isn't so good are reading handwritten documents. It just returns nothing instead. So I have an alternate route here to show how to upload the PDF to OpenAI and have gpt-5-mini read the document instead, which it can do with no problems. This workflow is more specific to Monday.com but you can pull out the logic for uploading and calling OpenAI's API to read the files you've uploaded, and put those nodes into whatever automation you want. The workflow JSON is attached. Let me know if you have any questions!

This is Part 3 in Automating YouTube video to Shorts using self-hosted n8n, Whisper, FFmpeg, and ChatGPT-4o. Finish up the series I'll walk through combining all the clips from Part 2 into one full short, and wrap up that short with some automatic subtitles.

With one HTTP Request node and a few regular nodes, you can pull images out of Monday.com and send them to GPT for analysis, like if you wanted to get data from receipts (@Jason Guan 😉) or transfer data from a PDF to a spreadsheet. Getting the files or images out of Monday.com is not the most straightforward process so I have included the workflow and the JSON below. JSON body to request the Monday.com Public URL for an image: { "query": "{ assets(ids: [{{ $json.assetId }}]) { id name public_url url } }" } The sample GPT prompt is attached as a separate file. Let me know if you need any help!