In this chapter, you learn about another revolutionary deep learning architecture, the variational autoencoder. Variational autoencoders (VAEs) are similar to autoencoders and GANs, learning the latent variables of the data to understand how to represent and create new data. Variational autoencoders have gained popularity in recent years due to their ability to describe observations in a probabilistic manner. This ability makes variational autoencoders able to be employed in a wide range of applications, for example, generating audio, video or text; anomaly detection; and denoising; just to name a few. In this chapter, you learn the theory behind the variational autoencoders and implement some examples using Julia and Python.