Suno vs MusicLM

Suno is revolutionizing the world of audio by generating realistic sounds, encompassing everything from speech to music and sound effects. While Bark was initially optimized to create lifelike speech, the same foundational model has the versatility to produce non-speech audio, including captivating music and immersive sound effects.

Excitingly, we are now previewing our innovative music model, Chirp, exclusively on Discord. By joining the server, you gain access to the new /chirp command, allowing you to create music with vocals - all for free and available on demand. Experience the future of audio creation with Suno!

Suno V3

Suno introduces v3, their latest musical innovation designed to empower anyone to create music. With v3, you can effortlessly generate complete, two-minute songs within seconds, no matter your musical experience. Whether you're an award-winning artist or someone experimenting with music creation for the first time, v3 caters to all by supporting song creation in any major language with just a few words.

The groundbreaking v3 model boasts significant enhancements in audio quality, a wider array of styles and genres, and improved prompt adherence, ensuring that your songs have fewer unintended quirks and smoother conclusions. Leveraging user feedback from the Pro and Premier members who beta-tested v3 Alpha, Suno has refined their product, addressing user needs and pushing the boundaries of AI-generated music. The platform also places an emphasis on trust and safety with proprietary watermarking technology that identifies music made with Suno to prevent misuse.

As Suno looks ahead, v4 is in the pipeline, promising new exciting features that continue the journey towards high-quality, controllable, and rapid music creation. Users are encouraged to share their creations on social media platforms such as X, Instagram, and TikTok using the handle @suno_ai_. Experience the future of music creation with Suno v3, available now for all users.

Google introduce MusicLM, a model generating high-fidelity music from text descriptions such as "a calming violin melody backed by a distorted guitar riff".

MusicLM casts the process of conditional music generation as a hierarchical sequence-to-sequence modeling task, and it generates music at 24 kHz that remains consistent over several minutes.