Automated Video Creation Workflow Using n8n

This workflow automates the creation and publishing of animated videos based on ideas listed in a Google Sheet. It processes one idea at a time, generating text prompts, images, animations, sound effects, and merging them into a final video before uploading it to YouTube.

Table of Contents

• Pre-conditions and Requirements
• Google Sheets Setup
• Step-by-Step Workflow Explanation
• Customization Guide

Pre-conditions and Requirements

1. API Keys Required

To run this workflow, you'll need API access to the following services:

• Anthropic Claude or Google Gemini (for text prompt generation)
• Flux AI (RapidAPI) (for AI-generated images) or
• Flux AI (BlackForestLabs) (I have included all image generation nodes for you to select the one you want) *Scroll to bottom to see new image generation reference from latest update
• RunwayML (API Documentation) (for AI video animation)
• ElevenLabs (for AI-generated voiceovers and sound effects)
• Creatomate (Website) (for video/audio merging and rendering)
• YouTube API (n8n docs)(for video upload and posting)
• New!! Kraken.io upload image to get url guide at the bottom.

2. n8n Instance Setup

• Sign up to n8n cloud or
• Install and run n8n (Official Guide)
• Set up credentials for each API in n8n’s settings

Google Sheets Setup

Here is the google sheet template for you to copy and get started with: https://docs.google.com/spreadsheets/d/1bq0opmTiFoP57JCtFRfHYfXwDvocbmC_JSaYgtBkung/edit?gid=0#gid=0 (click file and select make a copy so that you can have the sheet on your account)

Step-by-Step Workflow Explanation

1. Extract an Idea from Google Sheets

• The workflow retrieves the first row where videoStatus = "To Do".
• Marks it as Processing to avoid duplicate processing.

2. Generate a Prompt for AI Image Creation

• Uses Anthropic Claude or Google Gemini to generate a detailed prompt describing the scene.

3. Generate an AI Image with Flux AI

• Sends the prompt to Flux AI to create a high-quality image.

4. Animate the Image Using RunwayML

• The generated image is sent to RunwayML, which applies animation effects.

5. Generate Sound Effects and Voiceover with ElevenLabs

• ElevenLabs produces a realistic narration based on the video content.
• Background sound effects (e.g., storm sounds, fire crackling) are also generated.

6. Merge Video and Audio Using Creatomate

• Creatomate compiles the animated video with the audio.

7. Upload and Publish to YouTube

• The finalized video is automatically uploaded to YouTube using the YouTube API.

8. Update Google Sheet Status

• Marks videoStatus as Created.
• Marks publishStatus as Processed.

Customization Guide

Modifying the Animation Style

• Update the style column in Google Sheets with custom animation preferences (e.g., cinematic, slow-motion).
• Modify the prompt generation step in n8n to incorporate different styles.

Changing the Video Length

• Adjust the RunwayML settings to control animation speed and length.
• Make sure you also increase the wait time in video generation flow when you increase the length of you video, to make sure get url works. You have an option to add a
• Modify the Creatomate rendering step to adjust clip duration.

Adding Subtitles

• Modify the Creatomate step to include AI-generated subtitles from ElevenLabs' text output.

Publishing to Other Platforms

• Add additional steps to post to TikTok, Instagram, or Facebook using their respective APIs.

This workflow ensures a fully automated video creation pipeline, reducing manual effort and optimizing content production. 🚀

Image Generation Update Reference

For image generation in this workflow, I have provided two options, you can use flux using black forest labs or flux using rapidapi (has free playground trial).

All these options end with a get image node to get the n8n binary file.

After that we need to get an online image url for that binary file so we will use kraken.io for this.

Guide for setting up kraken node

1. create a free account here and make sure you confirm email on your inbox

2. get your api key and secret key here

3. Replace the placeholders in the body parameter; {"auth":{"api_key": "your_api_key", "api_secret": "your_api_secret"}, "wait":true}

leave the double quotes, just replace the placeholder text "your_api_key and "your_api_secret".

4. Thats it, now everything should work without an issue.

Note that you can just disconnect any image generation workflow with one you prefer, just unlink from first image generation node to get image node, do not disconnect the kraken from eleven labs.

And also make sure after plugging in a separate image generation flow, update the get image node appropriately to make sure you are grabbing the correct url expression.

On get image url for flux flow, I have used generic header auth. Here is the doc to set yours correctly https://api.us1.bfl.ai/scalar