Discover Suno AI, an AI capable of generating music in 30 seconds from a simple text request. The latest version even allows you to transform any text into a song. The result is stunning !
Artificial intelligence continues to develop, to the point that we wonder where and when it will stop! ChatGPT and its variations are already capable of generating all kinds of text and code, Midjourney and company create tailor-made images, while, a few days ago, OpenAI unveiled Sora, its new AI capable of generating videos of a breathtaking realism from a simple textual description (see our article). In short, it seems that artificial intelligence knows how to do everything, and it’s not Suno AI that is going to make us say the opposite!
This young American company has developed an AI of the same name capable of generating music with voices from a simple prompt. It has also just released version 3 of the tool in alpha, for the moment exclusively reserved for those who have subscribed to a paid subscription. Suno AI published a few extracts on its X account (formerly Twitter), and the result is simply stunning! Metal, new wave, electro, acid jazz… There’s something for everyone! We’ll let you judge for yourself!
Instrumentals, from mind to music. What’s playing in the background of your life?
Some songs we had fun making this week at the office:
Prompt: 80s atmospheric new wave
(Made by Helena, Social Media @ Suno) pic.twitter.com/40EZgaunyx— Suno (@suno_ai_) February 26, 2024
Suno AI: the AI with a musical ear
How Suno works is very simple, since you just have to write the instructions in natural language in the input field provided for this purpose: musical style, desired emotion, theme of the song, etc. A few seconds later, the tool generates a musical composition, which can contain instrumental music, but also sung lyrics. Once the music is created, you can then share it, download it or make changes to it.
Thanks to the “personalized” mode, it is possible to give the AI a text to sing, then indicate a musical style to create a composition. Some smart guys tried it with a few speeches from our politicians, and the result is absolutely hilarious! For example, we have a metal version of the famous “The Republic is me!” by Jean-Luc Mélenchon, or even a gangsta rap and jazz version of the speech “I, President of the Republic” by François Hollande.
Suno AI: the thorny problem of copyright
Version 3 of Suno brings its share of new features. It improves sound quality, generates songs faster, better supports languages other than English, and allows songs composed of two minutes to be extended – in addition to the initial minute and a half. The main new feature is the possibility of creating instrumental pieces, which makes it possible to generate a multitude of ambient music. For now, only those who have a paid subscription can take advantage of it.
You can go to the official website to listen to user creations. Free registration allows you to benefit from 50 credits, which is equivalent to creating 10 songs. Then you can switch to two paid offers: one at $10 per month for 2,500 credits (500 songs) and one at $30 per month for 10,000 credits (2,000 songs), in addition to monetization of creations and ‘a priority generation.
While Suno is undoubtedly an incredible tool, AI nevertheless raises the issue of copyright. Indeed, the company is hardly transparent about the database used to train its machine. We can fear that, like OpenAI with ChatGPT, it has done so with content protected by copyright… Furthermore, this innovation worries musicians who provide sound banks – which are used in particular in the world of audiovisual, advertising or documentary in order to illustrate sequences. Because Sino, with its very good sound quality, its rapid execution and its cost of a few dollars per month, could seriously overshadow them…