AI music generation platform Suno has officially launched its new generation of model family - Suno v6. This is the largest upgrade Suno has ever made, and also the first model series co-developed with music industry partners such as Warner Music Group, BMG, and Believe. The official said that the v6 model is faster, more expressive, with higher audio quality, and gives creators unprecedented precision control. Previous versions will be gradually retired.
Three Model Matrixes to Meet Different Creative Needs
Suno v6 is not a single model, but includes three versions:
- v6: The flagship model, aimed at Pro and Premier subscribers. It is stable, reliable, and precise, capable of consistently producing high-quality, professional music in various styles and genres. It is suitable for creators with clear goals.
- v6-wild: Also aimed at paying users, it is designed for exploration. It produces more random and varied outputs, often bringing unexpected, rich, and bold results, suitable for creators seeking inspiration or breaking the norm.
- v6-mini: An efficient version for all users (including free users). It generates faster and uses fewer resources. The official claims that its performance is better than any other free music generation model currently available, making it easier to turn ideas into complete songs quickly.
Core New Features: Multimodal Input + Precise Local Editing
One of the biggest highlights of v6 is support for multimodal creation. Users are no longer limited to text prompts; they can directly upload images, videos, or voice memos, and the system will convert them into music. For example, based on the atmosphere of an image, the emotion of a video, or a humming and description in a voice memo, it can generate songs in the corresponding style.
At the same time, creators can perform precise local modifications on already generated songs without regenerating the entire piece:
- Use natural language instructions to modify a specific section (e.g., "Change the chorus to be sung by a gospel choir"), while keeping the rest unchanged.
- Modify a single line or even a single lyric, keeping the rest intact.
- Extract elements from multiple existing works for remixing (for example, extract vocals from song A, beats from song B, then add new lyrics and specify an 80s synthwave style).
- Sample a specific time point riff, separate guitar or other instruments, and build a new beat around it.
In addition, the model's understanding of music language has significantly improved, allowing it to better grasp vocals, instruments, structure, emotion, reference style, and overall atmosphere, making complex creativity easier to realize.
Industry Collaboration and Future Directions
Suno emphasized that v6 is a new generation model trained from scratch, using authorized data from partners and user data, as well as integrating the team's previous technical and preference learning achievements. The platform still prohibits directly prompting the generation of imitations of specific famous artists or songs. The official also revealed that future plans include exploring new experiences such as AI remixes where artists voluntarily participate, and creating new revenue opportunities for rights holders.
Currently, Suno v6 has begun global distribution. Creators can experience the three models directly on the platform, achieving a complete loop from inspiration to finished product in less time. AIbase will continue to monitor subsequent updates and user feedback.
