Create ANYTHING with Sora 2 + n8n AI Agents In 10 Minutes

Key Concepts

• Sora 2: An advanced AI model for video generation.
• NADN: A platform for building and automating workflows, including AI integrations.
• Scenario: A workflow or automation sequence within NADN.
• Trigger: An event that initiates a NADN scenario (e.g., form submission).
• Form Submission: A trigger where user input from a form starts a workflow.
• AI Agent: A component within NADN that uses AI models (like OpenAI) to perform tasks, such as prompt generation.
• Edit Image Node: A NADN tool for manipulating images, including resizing.
• Merge Node: A NADN tool to combine different workflow paths.
• Generate Video Node (Sora 2): The NADN node that interfaces with the Sora 2 model for video creation.
• Prompt Engineering: The process of crafting effective text inputs for AI models to achieve desired outputs.
• UGC (User-Generated Content): Content created by users, often used in marketing.
• API Key: A credential used to authenticate access to services like OpenAI.
• System Message: Instructions given to an AI agent to define its role and behavior.

Sora 2 Video Generation Workflow in NADN

This summary details a step-by-step process for creating product advertisements using Sora 2 within the NADN platform. The workflow leverages user-provided descriptions and images to generate short, engaging video ads.

1. Setting Up NADN for Sora 2

• Accessing Sora 2: To use Sora 2 within NADN, users must ensure their NADN instance is updated. This is checked by looking for the "generate video" option.
• Updating NADN: If Sora 2 is not visible, users need to access the admin panel (cloud icon), navigate to "manage," select the "latest NADN version," and click "save changes." This update process takes approximately two minutes.

2. Initiating the Workflow: Form Submission Trigger

• Scenario Creation: The process begins by creating a new scenario in NADN.
• Trigger Configuration: The chosen trigger is a "form submission," which initiates the video ad creation workflow.
• Form Design: The form is titled "Sora 2 Video Generator" and includes two essential fields:
  • Video Description: A text field where users describe the desired video content.
  • Image Upload: A file field for uploading a reference image.
• Input Requirements: The form submission captures the video description and the uploaded image, which are the core inputs for Sora 2.
• Image Recommendations:
  • File Size: Images should ideally be under 2 megabytes to avoid potential issues with Sora 2 in its current stage.
  • File Format: JPEG is recommended for easier processing.

3. Image Preprocessing: Edit Image Node

• Purpose: Sora 2 requires specific image dimensions for video generation. The "Edit Image" node is used to resize the uploaded image to meet these requirements.
• Resizing Options: The Sora 2 model supports specific dimensions:
  • 720 x 1280 (suitable for vertical/mobile formats)
  • 1280 x 720 (suitable for horizontal/desktop formats)
• Configuration:
  • The user selects "resize" and inputs the desired width and height (e.g., 1280x720 for a desktop-oriented ad).
  • The "property name" is set to "image" to match the input from the form.
  • Crucially, the "ignore aspect ratio" option is selected. This ensures the image conforms to the target dimensions, even if its original aspect ratio differs, preventing potential compatibility issues.

4. Prompt Enhancement: AI Agent

• Objective: To automatically generate a high-quality, optimized prompt for Sora 2, saving users time and effort.
• AI Agent Setup: An AI agent is created within NADN.
• Prompt Generation Logic:
  • Input: The AI agent receives the "video description" from the form submission.
  • System Message: A system message defines the AI agent's role: "You're an expert at writing cinematic high converting UGC video prompts for Sora 2. Your job is to take the user's written concept and turn it into a single production ready prompt that transformed an uploaded image into a visual visually realistic short ad."
  • Rules: Additional rules are provided to guide the prompt generation for optimal results.
  • Model Selection: The "OpenAI chat model" is chosen for this task.
  • API Key Integration: To use OpenAI, an API key is required. This involves:
    • Navigating to platform.openai.com.
    • Signing in or creating an OpenAI account.
    • Verifying identity (a one-time requirement for newer models).
    • Going to "settings" and then "API keys."
    • Creating a "secret key" (e.g., named "test").
    • Copying the generated key and pasting it into NADN.
• Execution: The AI agent processes the input and generates an enhanced prompt.

5. Merging Workflow Paths: Merge Node

• Purpose: To combine the processed image data and the generated prompt into a single data stream for the Sora 2 node.
• Configuration:
  • A "merge" node is added.
  • The "combine" option is selected.
  • The "position" setting is used to merge the two input paths.
  • The output from the AI agent (containing the prompt) and the resized image are fed into the merge node.
• Outcome: The merge node consolidates the data, ensuring that both the image and the prompt are available for the subsequent video generation step.

6. Video Generation: Sora 2 Node

• Integration: The "generate video" node, which interfaces with Sora 2, is added to the workflow.
• Account Integration: The previously configured OpenAI API key is used to authenticate the connection.
• Model Selection: Users can choose between "Sora 2 Pro" and "Sora 2." The example uses "Sora 2" for cost-effectiveness.
• Video Parameters:
  • Duration: Users can specify the video length (e.g., 4, 10, or 8 seconds).
  • Dimensions: Users select the desired video dimensions (e.g., horizontal 1280x720).
  • Reference Image: The "reference" parameter must match the image input, which is set to "image."
• Prompt Input: The prompt is not directly entered but is dynamically pulled from the output of the merge node by selecting "schema" and then dragging the prompt from the merge output.
• Execution: The "execute step" command initiates the Sora 2 video generation process.

7. Real-World Applications and Examples

The transcript showcases the potential of this workflow through several examples:

• Crest Toothpaste Ad:
  • Input Image: Crest toothpaste tube.
  • Input Description: "This is an ad for Crest toothpaste."
  • Generated Video Output: A short commercial with a voiceover: "Just finished brushing with this and wow, my mouth feels so fresh, like squeaky clean fresh. If you want that just from the dentist vibe every morning, this is it. Like, I swear this is the exact kind of ad you'd see on television. It's crazy."
• Watch Ad:
  • Generated Video Output: A sophisticated ad highlighting precision and minimalism: "With precision, designed to endure a tiny piece that moves as effortlessly as you do. The essence of minimalism, elevated."
• Running Shoes Ad:
  • Input: A single running shoe image.
  • Generated Video Output: An atmospheric ad: "Are quieter before the city wakes up. It's just me, my breath, and a pair of shoes that feel like nothing at all. lightweight, supportive, ready to move with me."

8. Key Arguments and Perspectives

• Automation Potential: The primary argument is that NADN, combined with Sora 2, enables the automated creation of high-quality video ads, significantly reducing manual effort and time.
• Accessibility: The workflow is designed to be accessible even to users with no coding experience, with free blueprints provided.
• Efficiency: The AI agent for prompt generation and the merge node streamline the process, making it efficient.
• Versatility: The technology can be applied to various product types and marketing needs.
• Limitations: The transcript acknowledges that the generated videos are "not 100% perfect" but are "really, really cool and fast."

9. Notable Quotes

• "So, Sora 2 has been officially released inside Naden and the results are absolutely insane."
• "This is the exact kind of ad you'd see on television. It's crazy."
• "The blueprints are going to be 100% down below for free."
• "You're an expert at writing cinematic high converting UGC video prompts for Sora 2."
• "It's just me, my breath, and a pair of shoes that feel like nothing at all."

10. Further Resources and Community

The video concludes by directing viewers to additional resources:

• Social Media Automation: A separate video on automatically posting ads across platforms like X, LinkedIn, Instagram, and TikTok.
• NADN School Community: A community offering three main outcomes:
  1. Learning NADN and AI tools for time-saving and increased income.
  2. Creating an AI agency, with guidance on finding, closing, and fulfilling deals.
  3. Automating work for existing businesses, with blueprints for scaling to seven figures and automating 80% of operations.

Synthesis/Conclusion

The video demonstrates a powerful and accessible method for generating product advertisements using Sora 2 within the NADN automation platform. By combining a user-friendly form submission trigger, image resizing, AI-driven prompt engineering, and efficient workflow merging, users can create engaging video content with minimal technical expertise. The provided examples highlight the impressive quality and speed of this automated ad creation process, positioning it as a valuable tool for marketers and businesses looking to enhance their social media presence and streamline their content production. The emphasis on free blueprints and a supportive community further underscores the accessibility and potential impact of this technology.