Idea → Talking Head

What you’ll build

Type a topic — get a video of you talking about it in your voice. Every step runs locally on your own GPU. No subscription. No per-render meter. The chain is yours, the file is yours, the render lives on your drive forever.

The chain

4 pipeline nodes, in order:

1. 1
   Flux + Hybrig LoRALocal~30s – 1 min

   Renders a still photo of you using your trained LoRA.

2. 2
   Wan 2.2 I2VLocal~3 min – 4 min

   Animates a still photo into a short video clip.

3. 3
   F5-TTS (chunked)Local~8s – 20s

   Reads your script in your cloned voice.

4. 4
   LatentSyncLocal~20s – 45s

   Syncs the on-screen mouth to the spoken audio.

Prerequisites

• An NVIDIA RTX GPU with 12 GB+ VRAM. 24 GB recommended for the full pipeline. Check system requirements →
• Hybrig desktop installed and your worker connected. Get the desktop app →
• A trained character LoRA. Train one from 4–10 reference photos at /characters.
• A 10–30 second clean voice sample (your voice, recorded quietly). F5-TTS clones from the reference clip locally — audio never leaves your machine.

Try it

The fastest way to learn this chain is to drop it onto the Studio canvas and run it on your own rig. The whole graph is pre-wired.

Watch