AI-Powered Video Content Analysis with Gemini - n8n Workflow

Content Marketers: Automating the creation of detailed descriptions and tags for video assets.
Media Analysts: Quickly generating transcripts or visual descriptions for large video libraries.
Developers: Learning how to handle multi-part binary data uploads and complex API interactions within an n8n workflow.
SEO Specialists: Creating rich metadata for video content using AI-generated descriptions.

Analyzing video content manually is time-consuming and inefficient. This sophisticated n8n workflow solves this by fully automating the process using Google's powerful Gemini AI. The solution handles the entire video lifecycle: fetching the video from a URL, uploading the large binary file directly to the Gemini service using specific HTTP Request header commands, and finally prompting the AI model (Gemini 2.0 Flash) to deliver a comprehensive visual analysis. This n8n workflow is an excellent example of using custom HTTP Request n8n node configurations for advanced integrations where no dedicated node exists, providing a robust foundation for building high-performance aiautomationworkflows.

This n8n workflow operates in a distinct sequence of steps to manage the video file and communicate with the Gemini API:

1. Trigger and Input: The n8n workflow starts with a Manual Trigger (for testing) that feeds into the Set Input n8n node, providing a sample video_url.


2. Download: The Download video n8n node uses an HTTP Request to fetch the video data from the URL, transforming it into binary data ready for the next step.


3. Upload: The Upload video Gemini n8n node is configured with precise HTTP headers (X-Goog-Upload-Command, X-Goog-Upload-Header-Content-Length, Content-Type) to correctly stream the binary data to Gemini's file upload endpoint. This requires referencing a secure GeminiKey environment variable.


4. Flow Control: A Wait n8n node acts as a temporary delay or flow control mechanism, ensuring the upload process is registered before moving to the computationally intensive analysis phase.


5. Analysis Request: The Analyze video Gemini n8n node sends a second HTTP Request. This request uses the file URI returned from the previous upload step and includes a detailed system prompt asking the Gemini 2.0 Flash model to describe all visual elements, actions, branding, style, and tone of the video.


6. Output Extraction: Finally, the Get Result n8n node extracts the AI's detailed textual description from the complex Gemini JSON response, setting the final output variable videoDescription. This complete sequence demonstrates a high-utility n8n workflow for media processing.

To deploy and run this powerful n8n templates workflow, follow these steps:

1. Import: Copy the provided JSON data and paste it into your n8n instance using the 'Import Workflow' function.


2. API Key Setup: This n8n workflow relies on the Gemini API. You must set an environment variable named GeminiKey in your n8n environment or configure a secure Credential.


3. URL Modification: Locate the Set Input n8n node and replace the placeholder video URL with the URL of the video you wish to analyze. For a production environment, you would replace the initial manual n8n trigger with a Webhook or a monitoring n8n trigger like an S3 or Google Drive watch node.


4. Testing: Click the 'Test workflow' button on the manual n8n trigger to execute the flow and see the AI-generated analysis in the final 'Get Result' node.

When clicking ‘Test workflow’ (Manual Trigger): The starting n8n trigger for testing the flow and injecting the initial data.
Set Input (Set n8n node): Defines the video_url variable that the rest of the n8n workflow uses, streamlining input data management.
Download video (HTTP Request): Function is to fetch the video file content from the URL specified in the input, converting it into binary data ($binary.data) needed for the subsequent upload.
Upload video Gemini (HTTP Request): This complex n8n node handles the multipart file upload to the Gemini API. Key Configuration includes using the POST method, setting the contentType to binaryData, and defining crucial X-Goog-Upload headers to manage the upload command and file size.
Wait (Wait n8n node): Used here as a necessary flow control component to allow time for the potentially large video file to be processed by the Gemini servers post-upload.
Analyze video Gemini (HTTP Request): Sends the core analysis request. Key Configuration involves a detailed JSON body containing the uploaded file URI and a long text prompt asking the AI to describe the visuals, branding, style, and tone. This is the main intelligence component of the n8n workflow.

• Get Result (Set n8n node): Extracts the final analysis text from the API response payload into an easily consumable variable named videoDescription.