Mureka V6 - Kunlun Tech's AI music creation platform model

What is Mureka V6?

Mureka V6 is the foundational model of Kunlun Tech's AI music creation platform, supporting pure music generation and AI music creation in 10 languages. It incorporates self-developed ICL (in-context learning) technology, resulting in a wider soundstage and further enhanced vocal texture and mixing design. Users can create music using Mureka V6 in simple steps. The generated music is freely usable and suitable for various scenarios such as YouTube, advertising, and game background music. Mureka V6 supports multiple musical styles and emotional expressions, covering genres such as jazz, electronic, pop, and country.

Main features of Mureka V6

• Multilingual supportSupports AI music creation in English, Chinese, Japanese, Korean, French, Spanish, Portuguese, German, Italian, and Russian.
• Style controlUsers can specify the style of a song by uploading reference audio to ensure that the generated music matches the creative intent.
• Melody RecordingUsers can record melodic motifs, and the system will generate a complete musical accompaniment based on the recorded melody.
• Supports multiple styles and emotional expressionsIt encompasses a variety of musical styles, including jazz, electronic, pop, country, rhythm and blues, soul, blues, rock, and dance music, as well as a range of emotional expressions such as joy, indulgence, mystery, vitality, and sadness.
• Pure music generationUsers can generate instrumental music without needing to input lyrics.
• Personalized creationUsers can generate personalized music works through various input methods (such as text prompts, timbre references, and audio references).

Technical Principles of Mureka V6

• Self-developed ICL technologyMureka V6 introduces its self-developed ICL technology, enabling the model to better understand contextual information and generate music that better meets user needs.
• Structured generationUnlike traditional autoregressive models that generate audio incrementally, Mureka V6 pre-generates the overall music structure before predicting fine-grained audio tokens. This improves the structural coherence of the generated music and the accuracy of instrument arrangement.
• Diverse data sourcesThe training data for Mureka V6 includes synthetic data, human expert-annotated data, and the CoT Dataset. This diverse range of data sources provides the model with rich learning material, contributing to the generation of high-quality music.
• Reinforcement learning optimizationMureka V6 employs reinforcement learning techniques during training to optimize model performance through continuous validation and error correction.

How to use Mureka V6

• Visit the official websiteVisit the official Mureka website and click "Create" to enter the creation interface.
• Register an accountRegister using your email or social media account, or log in directly.
• Select creation modeAfter entering the creation interface, select "Simple Mode" or "Advanced Mode".
  • Simple modeEnter a song description of no more than 300 characters, such as "a song full of sincere emotions, expressing deep longing for friends", and then click "Create".
  • Advanced modeIt offers more creative options, such as uploading reference songs and selecting singers' voices.
• Select ModelDepending on your needs, click to select a model such as Mureka O1 or Mureka V6.

Application scenarios of Mureka V6

• Music loversMusic lovers can use Mureka V6 to transform scattered musical ideas into complete works, explore different musical styles, and improve their creative skills.
• professional musiciansProfessional music producers can use Mureka V6 to quickly create demos, saving early creative time, breaking through creative bottlenecks, and gaining new inspiration.
• AI streaming music platformUsers can input corresponding prompts based on the current scene or mood, and the platform will continuously generate customized music that matches the situation or emotion.
• Content creationMureka V6 can be used to generate custom background music for videos, podcasts, or other media projects, enhancing the appeal of the content.