Google DeepMind’s Lyria 3 model revolutionizes AI music generation by creating 30-second tracks with vocals, lyrics, and cover art based on simple text prompts or uploaded media. Users can input a genre, mood, or personal memory, and Lyria 3 produces a high-quality 48kHz stereo track. This model significantly improves on past versions by autonomously writing lyrics, allowing enhanced control over style, voice, and tempo, while also offering greater musical complexity. The feature is available on desktops now and will launch on mobile soon, supporting users 18 and older in eight languages, including German and English. Google ensures copyright compliance and originality, stating tracks are marked with a SynthID watermark. As AI music gains traction, with legal challenges from major labels, Google addresses these concerns while promoting Lyria 3 for creative expression. This innovation signals that AI music is evolving beyond novelty, potentially reshaping the industry landscape.
Source link
