Meta Platforms, Inc (NASDAQ: META) unveiled its latest AI innovation on Wednesday. AudioCraft, is a tool that enables users to generate high-quality audio and music simply from text prompts.
AudioCraft consists of three models: MusicGen, AudioGen, and EnCodec, Meta revealed in a blog post.
MusicGen, trained on Meta-owned licensed music, allows users to generate music from text prompts. AudioGen, on the other hand, leverages publicly available sound effects to produce audio based on textual cues. Meta also released an improved version of EnCodec, enhancing the quality of music generation by reducing "artifacts."
It's worth noting that Meta is open-sourcing MusicGen, AudioGen, and EnCodec, enabling researchers and practitioners to explore their own datasets and models.
The multinational founded by Mark Zuckerberg pointed out that, while generative AI has made significant strides in images, video, and text, audio generation has faced challenges due to the complexity of signals and patterns.
"The AudioCraft family of models are capable of producing high-quality audio with long-term consistency, and they're easy to use," the company wrote.
"We see the AudioCraft family of models as tools for musicians and sound designers to provide inspiration, help people quickly brainstorm and iterate on their compositions in new ways," Meta added. "We can't wait to see what people create with AudioCraft."