On September 4, Google opened up its music generation model to the general public and developers: complete songs in 44.1 kHz stereo, richer voices and arrangements, with no free tier.
So far limited to Flow Music, Google's music generation model is now accessible from Gemini, Google AI Studio, Google Vids and the Gemini API. Google presents Lyria 3.5 as its musical model with the best sound rendering to date, with more expressive voices and instrumental arrangements more provided than previous versions.
The version exhibited by the API produces whole pieces — verses, choruses, bridges — in 44,1 kHz stereo. Google has added a layer of support for the creation of a genre: choice or description of a genre, arbitration between vocal and instrumental rendering, and a series of ready-to-use templates for everyday uses such as ambient music, jingles or ringtones.
The chosen rate is charged to the unit produced, at $0.08 per complete song, without free level — a choice that contrasts with most Gemini models and reflects the cost of generating a long audio signal.
Read the full article on the original source
Go to the source →