Skip to content

NUWA

A research series of pre-trained visual synthesis models

Visit official site

What NUWA is

NUWA is described as a series of research projects developed by Microsoft, covering NUWA, NUWA-Infinity, NUWA-LIP, Learning 3D Photography Videos and NUWA-XL. The projects involve pre-trained models for visual synthesis that generate or manipulate visual data such as images and video.

What you can do with it

  • Generate and manipulate visual data
  • Pre-train models across multiple modalities
  • Carry out infinite visual synthesis
  • Repair images with language guidance

Model notes

  • Self-supervised learning of 3D photography videos is listed
  • Long video generation is listed among the features
  • A supplement calls the series a unified 3D Transformer pipeline for visual synthesis

Example uses

  • Generate new image or video content with NUWA models
  • Use NUWA-Infinity for infinite visual synthesis
  • Use NUWA-LIP for language-guided image repair

Pros & cons

✓ What we like

  • Several visual synthesis tasks are grouped in one series
  • Multimodal pre-training is listed among the features

! What to watch out for

  • The material presents it as a group of research projects
  • The listed tasks centre on generating or manipulating visual data

FAQ

Which projects are named?

The material lists NUWA, NUWA-Infinity, NUWA-LIP, Learning 3D Photography Videos and NUWA-XL.

What does the series work on?

The material describes pre-trained models for visual synthesis that generate or manipulate visual data such as images and video.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review