I Built a 24/7 Viral Shorts Machine with No-Code (free n8n template)

Automating Viral Shorts: A Five-Step AI System

Overview of the Automation Process

• The speaker introduces a fully automated system for creating viral shorts without coding, consisting of five simple steps.

• Step one involves extracting the main story from a Google Sheet and using an AI agent to generate prompts.

• In step two, high-quality images are created using Flux, followed by video creation in step three with Cling.

• Step four utilizes 11 Labs to produce unique sound effects for each video, while step five combines all elements into a final video using Creatmate.

• The speaker promises free access to resources needed for setting up this system quickly.

Workflow Demonstration

• Before testing the workflow, the speaker shows the Google Sheet template that organizes categories and brands as input data.

• The first test focuses on processing fast food chains, pulling relevant items marked as "to-do" from the sheet.

• The schedule trigger can be set to run multiple times daily; however, it processes only one item at a time to avoid duplication.

Data Processing Steps

• Brands are organized into an array for further manipulation; this allows splitting them into individual items for processing images, videos, and audio files separately.

• A prompt generator agent is employed next; it receives category and brand information to create tailored outputs for each brand involved in the process.

AI Agent Functionality

• Each brand's details (e.g., McDonald's, In-N-Out) are fed into user messages while system messages guide how the AI should behave during output generation.

• The overview of the agent emphasizes its role in transforming well-known brands into visually appealing digital content through detailed prompts across three media types: image, video, and audio.

Output Generation Mechanism

• To ensure structured outputs from the AI agent, specific output format requirements are activated. This leads to organized responses detailing brand names alongside their respective prompts.

• Open Router is utilized to access OpenAI's GPT 4.1 chat model for generating these creative outputs effectively.

Exploring FAL AI for Image and Video Generation

Overview of FAL AI Models

• FAL AI provides access to various text-to-image and image-to-video models, including options like image to 3D and text to video.

• The specific model being utilized for text-to-image generation is Flux Pro V1.1 Ultra, which allows users to input prompts in a playground environment.

Generating Images with Flux Pro

• Each request for an image costs approximately 6 cents, and users can test different prompts to see the generated results.

• To connect with the API, users must copy the URL, input their credentials, set content type, and provide a prompt.

Setting Up API Credentials

• Users need to create a new API key in FAL and save it as a header parameter named "authorization" for use in multiple nodes within their workflow.

• A second header must be added for content type set as application/json; this is crucial for sending requests correctly.

Sending Requests and Checking Status

• The body request includes fields such as "prompt" (for image generation), along with aspect ratio settings (9:6).

• After sending requests, users wait approximately 12 seconds before checking the status of the generated images using the request ID from earlier steps.

Downloading Generated Images

• Once all images are confirmed complete, they can be downloaded using the same request ID; this process ensures that all files are retrieved successfully.

• It’s important to re-enter credentials when checking status or downloading images to maintain secure access throughout the workflow.

Transitioning from Images to Videos

• After obtaining images, users will utilize FA's Cling model to convert these images into videos by providing both an image and a video prompt.

• Similar API setup procedures apply when working with Cling; endpoints must be configured properly along with necessary credentials.

How to Create Videos from Images and Add Sound Effects

Setting Up the Video Creation Process

• The process begins with sending a JSON request that includes a video prompt generated by an AI agent, which will be used for creating videos.

• An image URL is dragged into the workflow to convert images into videos, specifying that four different images are being processed simultaneously.

• After initiating the video creation, there is a waiting period of about 5 minutes to allow processing time before checking the status of the videos.

• The endpoint for fetching video status is referenced in documentation; it requires inputting a request ID to retrieve results.

• A polling method is employed to check if all videos are complete before proceeding to download them using their respective request IDs.

Reviewing Generated Videos

• Upon completion, examples of generated videos are reviewed, including one featuring Chick-fil-A and another showcasing Taco Bell.

• Each video created from images needs accompanying sound effects; thus, an HTTP request is made to 11 Labs for audio generation.

Configuring Audio Requests

• To create sound effects, an API key from 11 Labs is required. The correct header format must be followed when making requests.

• A body request containing text prompts for audio generation is sent; each audio file is set to be 5 seconds long, matching the length of the corresponding videos.

Making Audio Files Publicly Accessible

• Once audio files are generated, they need to be uploaded to Google Drive and made publicly accessible so they can be integrated with the video content later on.

• During upload, each audio file's name reflects its category (e.g., "fast food chains"), ensuring clarity in organization within Google Drive.

• After uploading, unique IDs for each audio file are obtained; these IDs will facilitate sharing permissions needed for public access.

Finalizing Audio Accessibility

• The next step involves sharing each uploaded audio file by its ID and granting read permissions so that anyone can access them. This ensures seamless integration with the final video product.

Video Creation Process Overview

Data Manipulation and Aggregation

• The process begins with data manipulation, where four video files and four audio files are set up for aggregation into a single clean item.

• The aggregated items include pairs of videos and audios (video one with audio one, etc.), which are essential for the next step in creating a video template.

Using Creatmate Template

• A specific template called "Famous Brands" is utilized, designed to be 20 seconds long with four 5-second audio and video files.

• To access the template, an API integration request is made; users can copy the necessary cURL command to set it up in their environment.

Setting Up API Integration

• Users must replace the template ID and API key in their setup to ensure proper functionality when sending requests to Creatmate.

• The system processes the request by confirming the selected template and elements, indicating its status as planned or processing.

Video Processing Completion

• After approximately 61.59 seconds of processing time, the final video is generated based on provided URLs.

• The completed video showcases dynamic visuals like moving cars and palm trees, emphasizing a West Coast vibe.

Updating Google Sheets

• Following video creation, an update operation is performed on Google Sheets to mark a specific row (e.g., "fast food chains") as done while uploading the final URL.

• This update ensures that relevant information about the completed task is accurately reflected in the spreadsheet.

Auto Uploading to Social Media Platforms

• The next phase involves auto-uploading content to platforms like Instagram, TikTok, and YouTube using Blotato's services.

• An HTTP request sends a JSON body containing the video's URL to Blotato for generating a new database URL for publishing across social media channels.

How to Post on Social Media Using Blot

Posting to Instagram

• To post on Instagram, a POST request is made to Blot with specific headers and a JSON body that includes the title "Fast Food Chains Reimagined" and the media URL.

• The Instagram account ID must be copied from the user's Blot account after logging in, which is essential for making the post.

Posting to TikTok

• Similar to Instagram, posting on TikTok involves a POST request with a slightly different JSON body; settings like "is AI generated" can be adjusted.

• Users need to log into TikTok, copy their account ID, and paste it into the appropriate field in the JSON body for successful posting.

Posting to YouTube

• For YouTube posts, both title and text fields are filled with "Fast Food Chains Reimagined," while additional settings include making the video unlisted and not notifying subscribers.

• Users must log into YouTube, copy their account ID, and insert it into the JSON body before posting.

Managing Posts

• Users can track posted content through their Blot account under 'published posts' where they can access links directly leading to each social media platform's post.

• If there are issues such as failed posts due to exceeding quotas or too many requests (especially noted with TikTok), users can check the 'failed posts' section for error details.

Setting Up Resources

• To set up this workflow in under 10 minutes, users should join a free school community linked in the description and download a provided JSON file associated with this video.

• After downloading, users import this JSON file into Naden workflows. A setup guide will assist them in configuring necessary API keys and other resources.

Workflow Configuration Steps

• Initial steps involve setting up a schedule trigger for how often posts should run and creating a Google Sheet template that needs copying into one's own Google environment.

• Connecting an AI chat model (like Open Router using GPT4.1), requires entering an API key. This step is crucial for automating responses or generating content based on user inputs.

Setup Guide for API Integration

Obtaining and Setting Up API Keys

• Users need to acquire their 11 Labs API key by visiting the 11 Labs website, as detailed in the setup guide.

• The source code required for Creatmate can be found in a PDF within the school post; users should copy this code to create a new template.

• After pasting the source code into Creatmate's source editor, users must replace the template ID with their own from Creatmate.

• Users are instructed to replace the placeholder API key in their project settings with their actual 11 Labs API key.

Integrating Botato for Automation

• To use Botato, users can sign up using code Nate30 for a discount and obtain an API key to integrate into their project.

• Once logged into Botato, users must link their social media account IDs to enable automated posting.

Cost Breakdown of Running the System

Overview of Costs Involved

• The chat model utilizes OpenAI through Open Router GBT4.1, costing less than a cent per run, making it economical.

• Image generation via Flux Pro costs approximately 6 cents per image; four images per run total around 24 cents.

Video Generation Expenses

• Video generation with Clingv 1.6 Pro is about $2 per short video (5 seconds), which is on the pricier side compared to other workflows.

Additional Monthly Subscriptions

• Sound effects from 11 Labs require a $5 monthly starter plan; Creatmate has a subscription fee of $41/month after a free trial period.

• Optional Botato subscription costs $30/month but offers discounts with promo codes; it facilitates automatic posting across multiple platforms.

Final Thoughts and Community Engagement

Encouragement for Viewer Interaction

• Viewers are encouraged to like the video if they found it helpful and consider joining a paid community focused on AI automations and learning opportunities.

Course Offerings

• Two main courses are available: "Agent Zero" covering LLM foundations and "10 hours in 10 seconds," focusing on designing time-saving automations.