Some applications now offer the ability to create music entirely through artificial intelligence. However, are these features truly reliable and do they actually produce high-quality work? The publisher's response is somewhat mixed. Suno, an artificial intelligence-based music creation platform, has introduced new features. Following music comes voice, and the tool now offers vocal synthesis. This feature is available in public beta on web and smartphone interfaces. This new feature, named "Speech," allows for the generation of synthetic voices from scripts or descriptions, creating music autonomously. The platform is gaining more features. According to Jack Brody, product manager at Suno, this new direction allows for the introduction of new features beyond music creation. Music will remain at the heart of Suno and our developments. However, our vision has always included other forms of human expression. The particularity of Speech is generating voice and music in a single coherent audio track. This approach distinguishes itself from existing tools by offering an integration between vocal and musical elements, although the user can disable the background music if they prefer a voice alone. Functionality and available options include two modes of use. The simple mode allows the desired content to be described by a simple text query. The advanced mode, on the other hand, offers more control by allowing the addition of a complete script. Additional settings also allow the adjustment of the voice gender, the speech style, and the level of variation from one generation to another. The maximum duration of generated content is about eight minutes. A product still under development, however, while the developers could confidently announce their tool, Suno acknowledges that Speech does not currently function perfectly. According to Brody, "imprecision can occur," notably British accents that "wander to Australia and return." The publisher is committed to gradually improving the function based on user feedback. In summary, this new feature, although very interesting, is not yet perfect.