Ask and you shall receive – in the form of an AI generated song produced by Google from your text prompts with their MusicFX tool.

Google have been upgrading their text-to-music tool MusicFX. Their experiment released last year and allows users to input a simple text prompt and get a completely original piece of music up to 70 seconds based on their prompt, created out of thin air.

Since launching the tool last year, over 10 million tracks have been created with MusicFX. Now, Google are building on its potential with brand new improvements to the text-to-music tool.

Google have expanded on their experimental tool to create higher-quality audio. They’ve also made the generation of music with the tool even faster. According to their tests, users are enjoying the results 3x more since their upgrades.

A screenshot of MusicFX featuring expressive chips, longer generation, and looping options.
Image credit: Google

Users of MusicFX can download the creations and share them around online. How does Google protect their generated content from being touted as someone’s original? Each creation is marked with a digital fingerprint using Google’s SynthID technology. These watermarks are imperceptible to the human ear but detectable by computers.

Alongside MusicFX, Google also announced updates to their image and writing generative tools. ImageFX has gained a new prompt interface allowing users to quickly experiment with a different style for the same generated image. TextFX, a tool for producing creative writing like lyrics and stories, has gained general improvements to user experience and navigation.

These Google Labs tools are currently available in the US, New Zealand, Kenya, and Australia for users to try out.