# What is Variational inference? Definition and meaning

Turns Bayesian inference into optimization by fitting a simple distribution to a complex posterior.

## Definition

**Variational inference**: Turns Bayesian inference into optimization by fitting a simple distribution to a complex posterior.

In short: Approximating hard posteriors with simpler ones.

## When it is used

VAEs, Bayesian neural networks and uncertainty estimation.

## Learn more

The full guide to **[Variational inference](/tools/variational-inference)** explains how it works, when to use it and how it compares with the alternatives, with examples. The [glossary](/glossary) lists every term on the site from A to Z.

Previous term: [VAE](/glossary/vae) · Next term: [Vertex AI](/glossary/vertex-ai)
