Full Setup Guide

How to Install SadTalker

GitHub, Local, Colab & pip — everything you need to get SadTalker running in minutes, on any platform.

Google Colab Local Install System Requirements Troubleshooting

Three Ways to Run SadTalker

Pick the method that fits your hardware and comfort level

Method	Best for	Setup time	GPU needed?
Google Colab	Trying it for free, no install	~5 min	No (Colab provides one)
Hugging Face Spaces	Instant testing in-browser	0 min	No
Local install (GitHub)	Repeated/private use, full control	20–40 min	Recommended

Option 1

Run SadTalker on Google Colab

No install required — free GPU included

Open the official SadTalker Colab notebook.

Go to Runtime → Change runtime type → GPU, then save.

Run the setup cell — this clones the repository and downloads the pre-trained checkpoints automatically. First run takes a few minutes.

Upload your source image and audio file in the designated cells.

Run the generation cell. Your output video appears inline and can be downloaded directly from Colab's file browser.

Tip: Colab's free tier disconnects after a period of inactivity, and GPU availability is not guaranteed — if you plan to generate several videos, a local install avoids queueing.

SadTalker running on Google Colab — notebook setup

Option 2

Local Installation from GitHub

Full control, no queue times — recommended for regular use

Clone the Repository

bash

git clone https://github.com/OpenTalker/SadTalker.git
cd SadTalker

Terminal output after cloning the SadTalker repository

Set Up a Python Environment

SadTalker requires Python 3.10.x specifically — newer versions frequently break dependency resolution.

With conda (recommended):

bash

conda create -n sadtalker python=3.10
conda activate sadtalker
pip install -r requirements.txt

Without conda (standard venv):

bash

python3.10 -m venv sadtalker-env
source sadtalker-env/bin/activate
pip install -r requirements.txt

Install FFmpeg

FFmpeg is required for video encoding and is not installed automatically.

Windows

Download from the official FFmpeg site and add it to your PATH.

macOS

brew install ffmpeg

Linux

sudo apt install ffmpeg

Download the Pre-Trained Checkpoints

The model weights are not bundled in the repo (they are several GB) and must be downloaded separately using the download script included in the repository, or manually from the checkpoints release page linked in the GitHub README. Place them in the checkpoints/ folder as instructed there.

Launch the WebUI

bash

python app.py

This starts a local Gradio server, usually accessible at:

url

http://127.0.0.1:7860

Open that address in your browser to use the same interface as the online demo, but running entirely on your own machine.

System Requirements

Minimum specs to run SadTalker, and what is recommended for best performance

Component	Minimum	Recommended
OS	Windows 10/11, macOS, Ubuntu Linux	—
Python	3.10.x	3.10.6
RAM	4 GB	16 GB+
GPU	None (CPU works, much slower)	NVIDIA RTX series with CUDA
Disk space	~5 GB (checkpoints + dependencies)	10 GB+

Troubleshooting Common Errors

Solutions to the most frequently encountered issues

Which Option Should You Choose?

Just testing it once?

Use Hugging Face Spaces or Colab — zero setup required.

Generating videos regularly?

Install locally — no queue times, no disconnects, full control over settings.

No GPU available?

Colab's free GPU tier is your best option; local CPU generation works but is significantly slower.