Get Started
Back to Templates

Create Deepfake Videos by Swapping Faces with Fal.ai Wan 2.2 and AWS S3

Created by

Created by: Sandeep Patharkar | ai-solutions.agency || sandy4v

Sandeep Patharkar | ai-solutions.agency

Last update

Last update 10 hours ago

Categories

Share

Animate Any Face into a Video with Fal.ai

Create stunning deepfake-style videos automatically by swapping a face from an image onto a source video.

This workflow provides a powerful, automated pipeline to perform video face-swapping using the Fal.ai API. It's designed to handle the entire asynchronous process: accepting a source video and a target face image, uploading them to cloud storage, initiating the AI job, polling for completion, and retrieving the final, rendered video.

Services Used Features
🤖 Fal.ai Leverages the powerful Wan 2.2 model for high-quality face animation.
☁️ AWS S3 Uses enterprise-grade cloud storage for reliable public file hosting.
🔄 Polling Loop Intelligently waits for the asynchronous AI job to complete before proceeding.
📥 n8n Form Trigger Provides a simple UI to upload your source image and video.

How It Works ⚙️

  1. 📥 Get User Input: The workflow starts when you upload a source video and a face image via the n8n Form Trigger.
  2. ☁️ Upload to Cloud: Both files are automatically uploaded to a specified AWS S3 bucket to generate the publicly accessible URLs required by the AI model.
  3. 🚀 Start AI Job: The public URLs for the video and image are sent in an HTTP Request to the Fal.ai API, which starts the asynchronous face animation process and returns a request_id.
  4. ⏳ Wait & Check: The workflow enters a polling loop. It Waits for one minute, then makes another HTTP Request to the Fal.ai status endpoint using the request_id.
  5. ✅ Check for Completion: An IF node checks if the job status is COMPLETED. If not, the workflow loops back to the Wait node.
  6. 🎬 Retrieve Final Video: Once the job is complete, the workflow makes a final HTTP Request to fetch the finished animated video.

🛠️ How to Set Up

  1. 🔑 Set Up Fal.ai Credentials: Get your API Key from Fal.ai. In n8n, go to Credentials, add a new Header Auth credential, and save your key. Connect this credential to all three HTTP Request nodes in the workflow.
  2. ☁️ Configure AWS S3: Add your AWS credentials in n8n. In the two AWS S3 nodes (Upload Video1 and Upload Image1), update the Bucket Name parameter to your own S3 bucket. Ensure your bucket permissions allow for public reads.
  3. ▶️ Activate and Run: Activate the workflow. Open the Form Trigger URL from the n8n editor, upload your files, and submit. The final video will be available in the execution log of the Get Final Video node.

Requirements

  • An active Fal.ai account and API key.
  • An AWS account with an S3 bucket configured for public access.
  • Alternative Storage: For a personal setup, you can replace the AWS S3 nodes with Cloudinary nodes. Just ensure the output is a public URL.

💬 Need Help or Want to Learn More?

Template Author: Sandeep Patharkar
Category: Content Generation / Content Marketing
Difficulty: Intermediate
Estimated Setup Time: ⏱️ 20 minutes