MiniMax Music 2.6 - MiniMax's brand-new AI music generation model

What is MiniMax Music 2.6

MiniMax Music 2.6 is a brand-newAI music generationmodel launched by MiniMax, achieving all-dimensional evolution from the underlying architecture to the creative experience. This version compresses first-packet latency to within 20 seconds, breakthrough-solving the "blind box" pain point, and achieves precise BPM locking, fine-grained section structure control, and deep emotional expression. In terms of acoustic quality, vocals are more natural and realistic, low-frequency performance is significantly enhanced, and guofeng timbres accurately restore the charm of traditional instruments. At the same time, it launches the Cover cross-style cover feature and the open-source AI Agent Music Skill ecosystem, supporting fun玩法 from melody transfer to pet singer.

Main features of MiniMax Music 2.6

• Intelligent music generation : Input a natural language description or lyrics, and AI automatically generates a complete song, supporting multiple genre styles (pop, rock, electronic, guofeng, etc.), and can precisely control BPM, tonality, and 14 section structures (such as [Intro], [Verse], [Chorus], [Bridge], etc.).
• Cover singing and remaking : After uploading reference audio, the model can extract the melody skeleton for cross-style adaptation (such as changing pop to jazz, folk to heavy metal), or retain the original melody and embed new lyrics to generate a cover work, realizing secondary music creation.
• AI Agent Music Skill (open-source ecosystem) Three skills are open-sourced for developers:
  • minimax-music-gen2: Agent automatically completes music generation through natural language descriptions
  • minimax-music-playlist: Automatically arranges personalized playlists based on user taste and scenarios
  • buddy-sings: Lets virtual characters improvise and sing in the first person, or transform pet sounds into singing
• Professional-grade audio control : Supports precise tempo locking, tonality control, and fine-grained section structure planning, solving the uncontrollable "blind box" problem of AI music and ensuring creative intent is faithfully restored.
• Multilingual vocal synthesis : Optimizes vocal performance, presenting a more natural and realistic singing feel, supports lyric generation in multiple languages such as Chinese and English, and in guofeng scenarios can precisely simulate the dynamic changes of erhu, dizi, guzheng, and opera singing styles.

How to use MiniMax Music 2.6

• Visit the official website and log in: Visit the official website of Music 2.6 https://www.minimaxi.com/audio/music，登录账号（新用户需注册）。当前处于14天免费内测期，C端用户每日可享500首免费创作额度。
• Choose a creation mode : On the main interface, choose the "Music Generation" mode to start creating. If you need to adapt an existing song, you can choose the "Cover" mode.
• Enter a creation prompt : In the input box, describe the music style, mood, scene, or purpose you want (such as "lyrical pop, a Chinese song about heartbreak, slow tempo"), or directly paste lyrics.
• Set music parameters(Optional): Set the BPM (tempo) and tonality (such as C major, A minor) according to your needs, and use structure tags to define song sections (such as [Intro][Verse][Chorus][Outro]) to control the musical structure.
• Generate and preview : Click the generate button, and initial audio can be obtained in about 20 seconds, and the effect can be previewed online.
• Fine-tune and adjust : If you are not satisfied with the result, you can adjust the prompt or parameters and regenerate, supporting multiple iterations of optimization.
• Download and share : Once satisfied, you can download high-quality audio files, or share directly to social platforms.
• Use the Cover feature(Advanced): For covers or style transfer, upload a reference audio, select the target style or fill in new lyrics, and a adapted version can be generated.
• API Access(Developer): Developers can obtain an API key through the MiniMax Open Platform and call the Music 2.6 interface to integrate it into their own applications. During the beta period, 100 free calls are granted daily.

Core Advantages of MiniMax Music 2.6

• Lightning-Fast Generation Experience : After a deep reconstruction of the underlying architecture, first-packet latency has been compressed towithin 20 seconds, allowing creators to receive initial audio feedback in just "the time of one deep breath" after entering a prompt, saying goodbye to long waits.
• Breakthrough in Precise Controllability : Completely solves the pain point of AI music being a "blind box" and achievesprecise locking of BPM and key、fine control of 14 section structures(such as [Verse], [Chorus], [Bridge], etc.), as well as deep emotional control, ensuring that creative intent is faithfully reproduced.
• Comprehensive Upgrade in Acoustic Quality
  • Vocal Naturalization: Sheds mechanical singing and presents a more casual and free real singing feel
  • Low-Frequency Enhancement: Deeper sub-bass extension and tighter cohesion, solving the muddiness problem in electronic music and game soundtracks
  • Special Optimization for Chinese Style: Accurately simulates erhu vibrato, dizi breath pauses, guzheng strumming, and dynamic changes in opera singing
• Industry-First Cover Feature : Supports cross-style transfer (such as pop to jazz, folk to heavy metal) and intelligent lyric-filled covers, providing new possibilities for secondary music creation.
• AI Agent Ecosystem Integration : Open-sourceminimax-music-gen2、minimax-music-playlist、buddy-singsthree major Music Skills, enabling AI Agents with capabilities such as music generation, playlist arrangement, and character singing.
• Deep Optimization for Chinese-Language Scenarios : Specially tuned for Chinese pop music, better fitting the aesthetic habits of Chinese users in melody lines, lyric articulation, and emotional expression.

Official Website Address of MiniMax Music 2.6

• Official Website Address：https://www.minimaxi.com/audio/music

Key Information and Usage Requirements of MiniMax Music 2.6

• Release Date: April 10, 2026
• Company: MiniMax (Xiyu Technology)
• Product Positioning: A new-generation AI music generation model, based on the abab 7 series large model architecture
• Version Features: A full-dimensional evolution version from the underlying engine to creative tools
• Generation Speed: First-packet latency <20 seconds
• Structure Control: Supports 14 section tags ([Intro], [Verse], [Chorus], [Bridge], [Hook], etc.)
• Musical Dimensions: Precise control of BPM (tempo) and key (C major, A minor, etc.)
• Timbre Performance: Vocal naturalization, low-frequency enhancement, and special optimization for Chinese-style instruments
• Cover Feature: Cross-style transfer and intelligent lyric-filled covers
• AI Agent Skill: Open-source three major skills: minimax-music-gen2, minimax-music-playlist, and buddy-sings
• Real-Time Generation: Returns preliminary audio results within 20 seconds
• Beta Period: 14-day free beta period (starting April 10, 2026)
• Consumer users: 500 free creations per day
• Developers: Additional 100 free API calls per day
• Exceeding limits: After the beta period ends, refer to official paid policies
• Prompts: Supports natural language descriptions (style, mood, scene) or directly pasting lyrics
• Structure tags: It is recommended to use tags such as [Verse], [Chorus] to clarify the section structure
• Parameter settings: BPM range and key can be specified to improve generation accuracy

Competitive analysis of similar products to MiniMax Music 2.6

Application scenarios of MiniMax Music 2.6

• Personal Music Creation : Ordinary users need no music foundation; they can create their own original songs simply by describing them in natural language, suitable for scenarios such as birthday wishes, emotional expression, and commemorative gifts, for example remaking a song from a family member's youth as a special gift.
• Gaming and Interactive Entertainment : Optimized for low- and mid-frequency performance specifically for game soundtracks, it can generate background music, battle BGM, and scene sound effects that match the game's atmosphere; developers can use the API to batch-generate audio assets adapted to the styles of different levels.
• Short Video and Social Media Content : Provides customized background music for short video creators, podcast hosts, and live streamers, supports rapid generation of soundtracks that match the emotional tone of the content, and solves the problem of limited copyrighted music.
• AI Agent and Intelligent Assistant Integration : Through the open-source Music Skill (minimax-music-gen2, minimax-music-playlist, buddy-sings), AI Agents gain capabilities such as music generation, personalized playlist arrangement, and virtual character singing, empowering applications such as intelligent customer service and virtual companionship.
• Guofeng and Traditional Culture Content : With specialized Guofeng optimization capabilities, it generates musical works that accurately simulate the erhu, dizi, guzheng, and operatic vocals for scenarios such as ancient-style videos, hanfu showcases, traditional culture promotion, and opera education.