VAEDecodeAudio - ComfyUI Built-in Node Documentation

The VAEDecodeAudio node converts latent representations back into audio waveforms using a Variational Autoencoder. It takes encoded audio samples and processes them through the VAE to reconstruct the original audio, applying normalization to ensure consistent output levels. The resulting audio is returned with a sample rate of 44100 Hz by default, or the sample rate from the input samples if provided.

Inputs

Parameter	Description	Data Type	Required	Range
`samples`	The encoded audio samples in latent space that will be decoded back to audio waveform	LATENT	Yes	-
`vae`	The Variational Autoencoder model used to decode the latent samples into audio	VAE	Yes	-

Outputs

Output Name	Description	Data Type
`AUDIO`	The decoded audio waveform with normalized volume and sample rate (default: 44100 Hz, or the sample rate from the input `samples` if present)	AUDIO

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): 2a3f5c912d1d84eea7768979f6b8f0eaa9fe89041f3a3352434f38abd3c09fea

​Inputs

​Outputs

Inputs

Outputs