Skip to content

FoleyCrafter

Generate semantically matched, time-aligned audio for video

Visit official site

What FoleyCrafter is

The framework is described as a text-based video-to-audio system that generates high-quality audio semantically related to the input video and aligned with it in time. It was developed by Shanghai AI Laboratory and the Chinese University of Hong Kong, Shenzhen.

What you can do with it

  • Generate high-quality audio semantically related to the video
  • Customise audio generation with text prompts
  • Align sound with visual cues in time
  • Use negative prompts to avoid unwanted audio effects

Application scenarios

  • Add ambient sound to a natural scenery video
  • Generate matching sound effects for an animation video
  • Add background music and ambient sound to a silent video

Getting started

  • Visit the official site
  • Read the product introduction and method notes
  • Browse the gallery for different video types
  • Download the code or try the online demo

Pros & cons

✓ What we like

  • Semantic relevance and timing sync are built into the output
  • Code, a demo and a gallery are available

! What to watch out for

  • Output follows the given video and text prompts
  • Negative prompts are used to steer away unwanted audio

FAQ

What does it generate from a video?

The material describes high-quality audio that is semantically related to the video and aligned in time.

Can the audio be steered?

Text prompts customise the output, and negative prompts avoid unwanted audio effects.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review