This is a comprehensive video that covers 5 ways (local and cloud based) to host n8n.

1. npm
2. docker
3. render
4. railway
5. digital ocean

I learned a lot making this video. Now I can save my clients' money. I am sure they will appreciate it !

Greetings everyone,

I have a few different sidestreams and I am currently using n8n at the core of my work from YouTube, Gumroad, to client work.

I find June's revenue to be a validation once again, especially for those interested in appearing in AI Search ($84 out of $110)

It helped me quite a lot that I my YouTube channel has grown. I added 83 subs. The video that helped me out the most was my n8n automation for creating Veo3 videos.

Keep making.

In the last post, I shared how you can self-host via npm. If you are looking to run n8n locally for yourself without overhead -- that's the best way!

However, if you are considering of having cron triggers or using n8n as your backend 24/7, then this is a great tutorial for you.

All in all, it's quite simple.

Steps

If you have downloaded Docker, go to Step 3

1. Download Docker from Docker dot com — pick the version that matches your computer
2. Install it > Open the Docker App on Desktop
3. Docker Hub > Search for n8n
4. Choose n8nio/n8n > Click “Pull”
5. Images > Click the “Run” button to start > Expand the dropdown menu of the Optional Setting
6. Ports: you can input whatever port you like, e.g. 5678, under the host port field > Run

Full YT Tutorial

A client of mine needs to cut down on costs. (So do I!)

Turns out, you can n8n locally for FREE.

It's as simple as:

• npx n8n
• n8n

Then you get : http://localhost:5678

Here are the full length tutorials on Medium and YouTube

I recently submitted a head to head comparison of the 3 major AI model providers using Lovable. I tested the same prompt and got interesting results.

Chose to follow up and repeat with Bolt, Replit, and Same.

Long story short -- they perform virtually identical to each other, because they all use Anthropic under the hood.

They all have different base prompts.

Check the full YT video here.

Summary

This last weekend Lovable hosted an impromptu hackathon called the AI Showdown.

I made an entry for the $25K competition.

Its rules are simple:

• Compare head to head the performance of OpenAI vs Anthropic vs Gemini.

So I decided to create a very simple submission that doubles down as an evaluation for future models and other platforms as well.

Namely, create a 2D Pokémon style game called Kanjemon, where the goal is to capture all 25 Kanji (Japanese scripture characters)

Motivation

I make language learning apps.

I also wanted a very simple comparison.

Results

Honestly, it's possible to 1-shot prompt an amazing MVP on Lovable -- no auth, just pure gameplay.

My personal favorite is Anthropic (better style taste), but OpenAI was incredible too (followed all instructions)

I wrote a lengthier write-up here, where I include the prompts and logic behind the submission.

YouTube demo

This flow was born out of the success of its predecessor -- a simple n8n automation that invoked the Fal AI API.

I wanted to go deeper in Veo3's capabilities.

I have to say it's great at generating ads.

Check this (Adidas) ad out !

In a nutshell the workflow is divided into three main phases: 

1. Create Veo3 Video 
2. Wait for Video Processing 
3. Post on YouTube 

Check out the full demo here.

Let me know if you build anything -- link it below !

I have been really active on YouTube in the last 2 months.

My niche is mostly automations using nocode tools, specifically n8n.

Yesterday, I got notice that Veo3 is accessible by API.

So I decided to make a video about : Veo3 Video creation using n8n

I posted on YouTube and Medium.

The result ?

Top of the ranks on Google Search for 'n8n veo3 tutorial'

This wouldn't have happened, if I didn't hustle for it.

Keep making.

Today I got notice that the Veo3 API is publicly available through fal ai.

So I made a video that walks you through.

It's an impressive model BUT quite expensive. It's ~1$ per video second.

Here's the demo!

A bit about the flow.

How It Works

1. A user submits a video prompt through a form.

2. An AI agent (using OpenRouter's Google Gemini 2.5 Flash model) refines the user's prompt into a detailed "script-to-screen" format suitable for video generation.

3. The refined prompt is sent to the FalAI Veo 3 model via an HTTP request to generate the video.

4. Details of the video request, including the date, the refined prompt, and the request URL, are stored in a Google Sheet.

Setup

To run this workflow, you need to set up credentials in n8n for:

* OpenRouter: Generate API key from your OpenRouter account.

* Fal AI: Generate API key from your FalAI account

* Google Sheets: Uses OAuth 2.0. Connect by authenticating your Google account.

Hey guys,

This tutorial walks you through along every step of the way. It's a tedious process, but you can do it.

Reason I made this is because WhatsApp has 1+ Billion users. Lots of clients have asked me to help them with a WhatsApp chatbot.

In summary what I did was:

Sett up the WhatsApp Trigger in n8n

Configure the Facebook/Meta Side

Connect n8n and Setting up the AI

Configure the WhatsApp Send Message Node

Finalize and Test the n8n Workflow

Activate and Live Test the Full System

Here's a full length write up on Medium.

Here's the YT Tutorial.