Gemina now lets you make music with simple prompts

Gemina has introduced a new feature that allows users to generate music using simple text or image prompts. This feature uses the latest music generator model of DeepMind called Lyria 3. The feature is currently available in beta and free, but Google AI Plus, Pro, and Ultra subscribers can benefit from a higher usage limit.

In line with its goals to develop creative tools based on artificial intelligence, Google continues to add new capabilities to Gemina. Thanks to the new music generation feature, songs can now be created with lyrics, vocals and wordless music. Simply describe an idea like a laid-back, humorous R&B track about a sock finding its match, and a quality track can be produced in seconds. Additionally, Gemina can be asked to draw inspiration from uploaded content to further personalize the music production process.

Google wrote in a blog post that the goal of these tracks is not to create a musical masterpiece, but to provide a fun and unique way to express yourself. The company also announced that all parts made in Gemina are equipped with SynthID; An invisible watermark to identify content generated by Google artificial intelligence.

Google has implemented a series of improvements in Lyria 3 compared to previous Lyria models. For example, users no longer need to write lyrics themselves, as lyrics are automatically generated based on prompts. The new model also offers more creative control, allowing users to choose elements such as style, vocals and tempo. In addition, Lyria 3 is capable of producing more realistic and musically complex tracks.

To use this new feature, you must first open the Geminai application or website. Then select Create music from the tools section and enter your desired description for the piece. This tool creates 30-second clips with custom cover art produced by Nano Banana. Finally, you can download the piece or create a shareable link to send to your friends.