AI Celebrity Face-Swap Ads
Budget: $30 – $250 USD
I have subscription to N8N and eleven labs and want to ability to create AI generated face swapping videos at low cost this is the suggested plan; I have a very low budget for this This is the worklow ; Summary of "Free" Workflow
n8n gets the script → ElevenLabs makes the .mp3.
n8n triggers FaceFusion CLI → Swaps face in the video.
n8n triggers Wav2Lip CLI → Syncs mouth to the .mp3.
To achieve this, you will use **n8n** as the orchestrator and **ElevenLabs** for the audio. Since you want to do this at **no cost** (avoiding expensive APIs like HeyGen or DeepSwap), you must run the video processing locally on your computer using an open-source tool called **FaceFusion**.
### 1. The Local Setup (One-Time)
Before n8n can "talk" to your computer's video processor, you need the engine installed.
1. **Download [Pinokio](https://pinokio.computer/):** This is a free browser that installs complex AI tools with one click.
2. **Install FaceFusion:** Inside Pinokio, search for "FaceFusion" and click install.
3. **Note the Path:** Note where FaceFusion is installed (usually `~/pinokio/bin/facefusion`).
### 2. The CLI Command
In n8n, you will use the **Execute Command** node. This node tells your computer to run FaceFusion in the background without you opening the app.
The "Universal" command for a face swap looks like this:
```bash
python run.py --source /path/to/your_face.jpg --target /path/to/viral_video.mp4 --output /path/to/final_result.mp4 --frame-processors face_swapper face_enhancer
```
* `face_swapper`: Swaps the face.
* `face_enhancer`: Cleans up the "blurriness" (this makes it look professional).
---
### 3. The n8n Workflow Strategy
You can copy-paste the logic below into a new n8n workflow. This flow takes text, generates a voice, and then triggers the local swap.
#### **Step 1: ElevenLabs Node**
* **Action:** Text to Speech.
* **Input:** Your script.
* **Output:** Binary file (audio).
#### **Step 2: Read Binary File Node**
* Point this to your "Viral Video" (the target) and your "Face Image" (the source) stored on your hard drive.
#### **Step 3: Execute Command Node**
This is the "Brain" node. Ensure your n8n settings allow the Execute Command node (if self-hosted).
**Command Template:**
```bash
# This navigates to your FaceFusion folder and runs the swap
cd /your/path/to/facefusion && python run.py --source {{ $json.face_path }} --target {{ $json.video_path }} --output /outputs/final_video.mp4 --frame-processors face_swapper face_enhancer --headless
```
*> **Note:** Using `--headless` is key; it prevents the FaceFusion window from popping up so the automation stays "silent."*
---
### 4. How to Lip-Sync (The Secret Sauce)
Face swapping doesn't move the mouth to the ElevenLabs audio. To do that for free, install **Wav2Lip** (also available in Pinokio).
**Add this second command to your n8n "Execute Command" node:**
```bash
python wav2lip.py --face /outputs/final_video.mp4 --audio /path/to/elevenlabs_audio.mp3 --outfile /outputs/VIRAL_READY.mp4
```
n8n gets the script → ElevenLabs makes the .mp3.
n8n triggers FaceFusion CLI → Swaps face in the video.
n8n triggers Wav2Lip CLI → Syncs mouth to the .mp3.
To achieve this, you will use **n8n** as the orchestrator and **ElevenLabs** for the audio. Since you want to do this at **no cost** (avoiding expensive APIs like HeyGen or DeepSwap), you must run the video processing locally on your computer using an open-source tool called **FaceFusion**.
### 1. The Local Setup (One-Time)
Before n8n can "talk" to your computer's video processor, you need the engine installed.
1. **Download [Pinokio](https://pinokio.computer/):** This is a free browser that installs complex AI tools with one click.
2. **Install FaceFusion:** Inside Pinokio, search for "FaceFusion" and click install.
3. **Note the Path:** Note where FaceFusion is installed (usually `~/pinokio/bin/facefusion`).
### 2. The CLI Command
In n8n, you will use the **Execute Command** node. This node tells your computer to run FaceFusion in the background without you opening the app.
The "Universal" command for a face swap looks like this:
```bash
python run.py --source /path/to/your_face.jpg --target /path/to/viral_video.mp4 --output /path/to/final_result.mp4 --frame-processors face_swapper face_enhancer
```
* `face_swapper`: Swaps the face.
* `face_enhancer`: Cleans up the "blurriness" (this makes it look professional).
---
### 3. The n8n Workflow Strategy
You can copy-paste the logic below into a new n8n workflow. This flow takes text, generates a voice, and then triggers the local swap.
#### **Step 1: ElevenLabs Node**
* **Action:** Text to Speech.
* **Input:** Your script.
* **Output:** Binary file (audio).
#### **Step 2: Read Binary File Node**
* Point this to your "Viral Video" (the target) and your "Face Image" (the source) stored on your hard drive.
#### **Step 3: Execute Command Node**
This is the "Brain" node. Ensure your n8n settings allow the Execute Command node (if self-hosted).
**Command Template:**
```bash
# This navigates to your FaceFusion folder and runs the swap
cd /your/path/to/facefusion && python run.py --source {{ $json.face_path }} --target {{ $json.video_path }} --output /outputs/final_video.mp4 --frame-processors face_swapper face_enhancer --headless
```
*> **Note:** Using `--headless` is key; it prevents the FaceFusion window from popping up so the automation stays "silent."*
---
### 4. How to Lip-Sync (The Secret Sauce)
Face swapping doesn't move the mouth to the ElevenLabs audio. To do that for free, install **Wav2Lip** (also available in Pinokio).
**Add this second command to your n8n "Execute Command" node:**
```bash
python wav2lip.py --face /outputs/final_video.mp4 --audio /path/to/elevenlabs_audio.mp3 --outfile /outputs/VIRAL_READY.mp4
```