以下文章来源于AGI接口 ，作者陈伊凡 出品｜虎嗅科技组 作者｜陈伊凡 编辑｜苗正卿 头图｜AI生成 "AI 原生 100" 是虎嗅科技组推出针对 AI 原生创新栏目，这是本系列的第「 15 」篇文章。 "我们确信不久的将来会有一个'大饼'掉下一个，虽然具体什么时候掉下来我们不知道，但我们要先 把盘子做好、做大，否则掉下来了我们也接不住。" 姜涛如此描述自己现在做的事。说这句话时，姜涛正坐在音潮办公室的会客区里。 AGI接口 . AI卷起的财富风暴。 那段时间，快手的办公室里总飘着奇奇怪怪的旋律。姜涛带着算法团队和阿怪挤在同一个隔间，一边 用代码训练模型，一边听阿怪讲和弦走向、编曲等乐理知识。这种奇妙的碰撞，成为了他音乐素养的 启蒙。 他亲历了中国AI音乐发展全过程，在过去的十多年时间，他几乎没有离开过AI音乐赛道，如今，他 是一家专注音乐大模型和AI音乐产品的公司——音潮的CEO。 如果第一次见到姜涛，很容易将他和音乐人联系在一起，但你很难想到这是一位从事了多年AI算法 的技术男。他穿着米色休闲亚麻西装，有时候还会带着一个圆形礼帽，很像一个玩爵士乐的"老炮 儿"。他会给妻子做歌，把女儿稚嫩的声音编进旋律，浪 ...

Investment Rating - The report maintains a "Buy" rating for NetEase Cloud Music (9899.HK) [4] Core Insights - The company reported a slight revenue miss but exceeded profit expectations due to a reduction in sales expenses. Revenue for 1H25 was RMB 38.3 billion, down 6% year-over-year, compared to Bloomberg's consensus estimate of RMB 39.1 billion. Gross profit reached RMB 13.9 billion, with a gross margin of 36.4%, an increase of 1.4 percentage points year-over-year. Operating profit was RMB 8.45 billion, up 40.8% year-over-year, primarily due to reduced sales expenses. Adjusted net profit attributable to shareholders was RMB 19.5 billion, a significant increase of 121% year-over-year, mainly due to the recognition of deferred tax assets amounting to RMB 8.5 billion [1][4] Revenue Breakdown - Online music service revenue for 1H25 was RMB 29.7 billion, reflecting a year-over-year growth of 15.9%. Subscription revenue grew by 15.2% to RMB 24.7 billion, driven by an increase in membership subscriptions, while social entertainment services and other revenues fell by 43.1% to RMB 8.6 billion due to the closure of the live streaming entry after the 2024 year-end version update [2][3] User Engagement and Product Development - The platform continues to enhance user engagement through product innovation and community ecosystem improvements. The NetEase Cloud Music app has introduced features such as personalized recommendations and AI functionalities, maintaining a daily active users to monthly active users ratio (DAU/MAU) above 30% [3] Cost Management - The company has effectively controlled costs, with sales expenses decreasing by 55.8% year-over-year to RMB 1.63 billion, attributed to a cautious promotional strategy. Management expenses increased by 3.6% to RMB 930 million, while R&D expenses decreased by 4.2% to RMB 3.79 billion [3] Profit Forecast and Valuation - The adjusted net profit forecast for 2025 has been raised to RMB 29.4 billion, a 55% increase from previous estimates. The adjusted net profit for 2026 and 2027 is also revised upward to RMB 25.9 billion and RMB 28.6 billion, representing increases of 19% and 17% respectively [4][5]

Core Viewpoint - The article discusses the launch of the "Yin Chao" music model and its new feature "Yin Guo," emphasizing its innovative capabilities in music creation and the potential to reshape the music industry landscape in China [2][4][5]. Group 1: "Yin Chao" Music Model - The "Yin Chao" music model has independently completed the WAIC 2025 English theme song "AI For Good," focusing on the theme of "technology for good" [2]. - It utilizes an industry-leading AR+NAR architecture, enabling high-fidelity reconstruction and capturing global dynamic changes [2]. - The model addresses traditional challenges in music detail preservation, generating high-quality music with rich arrangement layers [2]. Group 2: New Feature "Yin Guo" - The upcoming "Yin Guo" feature allows for multi-style fusion in music creation, aiming to surpass similar international products [4]. - It employs a self-developed contrastive learning model to blend various musical elements, enabling users to easily adjust the fusion ratio [4]. - Users can upload text, images, or videos as creative references, allowing the system to generate music that aligns with specific themes [4]. Group 3: Market Position and Vision - The CTO and CEO of the company stated that "Yin Guo" aims to provide domestic musicians with controllable, high-fidelity, and low-threshold generative tools, promoting a new paradigm in the local music industry [5]. - The introduction of "Yin Chao" signifies a shift towards democratizing music creation, breaking down resource monopolies, and fostering a win-win ecosystem for creators, users, and platforms [5][6]. - The model ensures that copyright benefits primarily belong to the creators, with a unique song creation certificate generated for each user [6]. Group 4: Cost Efficiency and Industry Impact - "Yin Chao" boasts significant advantages in generation efficiency and cost control, with the cost of generating a single song being less than one-fifth of current market alternatives [6]. - This innovative application is expected to ignite creativity among both casual users and professional musicians, reshaping production and consumption models in the music industry [6]. Group 5: Presentation Intelligence (PI) - In addition to "Yin Chao," the company has developed the AI-native presentation design intelligent agent, PI, which aims to redefine content creation paradigms [8]. - PI focuses on efficient collaboration between professional knowledge and design agents, providing a comprehensive system for multi-modal content creation [8][9]. - The platform supports dynamic balancing between human creativity and AI task scheduling, enhancing the overall user experience [10][11].

Core Viewpoint - The article discusses the rapid advancement of AI-generated music, particularly focusing on the capabilities of the new music model Mureka V7 developed by Kunlun Wanwei, which significantly surpasses its predecessors and competitors in various performance metrics [6][8][51]. Group 1: Mureka V7 Performance - Mureka V7 has been released as the strongest domestic music model, outperforming the overseas AI music platform Suno in key metrics such as average performance rating and overall audio quality [6][8]. - Compared to its predecessor Mureka V6, Mureka V7 shows substantial improvements in music quality, including melody and arrangement, as well as vocal and instrumental realism [7][8]. - The performance metrics for Mureka V7 include an average performance rating of 57.7%, mixing quality of 39.0%, and vocal realism of 70.0% [8]. Group 2: Features and Innovations - Mureka V7 introduces a feature allowing users to upload audio or video links to create songs mimicking specific artists, enhancing personalization in music creation [12][13]. - The model can analyze user-uploaded music to generate original works with similar styles, demonstrating its versatility in music generation [17]. - Mureka V7 has also upgraded its capabilities to generate music videos alongside audio, expanding its creative offerings [20]. Group 3: MusiCoT Technology - The MusiCoT technology has been optimized in Mureka V7, allowing for a structured approach to music creation that aligns with human creative processes [25][28]. - MusiCoT enables the model to generate music with clear structure and coherence, enhancing the overall quality of the output [29][33]. - The technology has shown superior performance in both subjective and objective evaluations, establishing a new standard in the industry [32][34]. Group 4: Voice Model Development - Kunlun Wanwei has also introduced Mureka TTS V1, an audio model that allows for customizable voice generation based on user-defined characteristics [39][40]. - This model surpasses competitors in various aspects of voice synthesis, indicating a strong position in the voice generation market [41]. - Mureka TTS V1 can create voices for various applications, including film, gaming, and advertising, broadening its market potential [45]. Group 5: Industry Trends - The article notes a shift in the industry towards the commercialization of AI models, with a focus on vertical models like music and video generation becoming the new competitive landscape [47][48]. - Kunlun Wanwei's strategy aligns with this trend, aiming to create a comprehensive ecosystem for AI-generated content across multiple domains [49][50]. - The growing user base of Mureka, with nearly 3 million new users since March, highlights its acceptance and impact on music creation [51].

Core Viewpoint - The article discusses the launch of Mureka v7, an AI music generation model, which is positioned as a competitive product in the domestic market, capable of producing high-quality music comparable to Suno 4.5 [1][11][14]. Group 1: Mureka v7 Overview - Mureka v7 is highlighted as one of the few AI music products in China, with a focus on its quality and user experience [1][11]. - The author emphasizes the ease of use and the ability to generate music by inputting song structures and lyrics [14][22]. Group 2: Song Structure and Creation - The article outlines the importance of song structure, detailing elements such as intro, verse, pre-chorus, chorus, instrumental break, bridge, and outro [23][24][25][26][27][28][29]. - It provides a formula for song structure, suggesting two main types: a simple structure and a more complex one that includes pre-choruses and bridges [32][34]. Group 3: Lyric Writing and AI Interaction - The author shares a template for generating lyrics, emphasizing the need for emotional depth and structural adherence to ensure AI can recognize and generate music effectively [36][50]. - The article suggests using external resources and AI tools to enhance the lyric writing process, indicating that Mureka can integrate with other platforms for better results [38][40]. Group 4: Copyright and Ownership - Mureka offers a significant advantage in terms of copyright, allowing users to download a certificate of ownership for their created music, contrasting with other platforms that may have restrictive policies [74][75][71]. - The article notes the evolution of AI music generation, highlighting Mureka's role in lowering the barriers for music creation [76][78].

Group 1 - The core concept of "A-Pop" is introduced by Timbaland, focusing on AI-driven music creation rather than national identity, with the first AI idol TaTa representing this new genre [3][7] - TaTa is developed by Stage Zero, a company co-founded by Timbaland and others, utilizing AI technology for music production, showcasing a shift towards fully autonomous AI artists [5][10] - The emergence of TaTa signifies a pivotal moment in the evolution of virtual idols, marking the transition to a new phase where AI can independently create and express [30][40] Group 2 - The global virtual idol market is rapidly expanding, with projections indicating it will exceed $5.03 billion by 2030, growing at a compound annual growth rate of 18.8% [34][36] - The Asia-Pacific region, particularly China, Japan, and South Korea, is leading the virtual idol market, highlighting the cultural significance and commercial potential of AI-driven entertainment [31][34] - Virtual idols are increasingly seen as controllable content assets, free from the limitations of human performers, thus reshaping the music industry and content creation processes [38][40] Group 3 - The evolution of virtual idols can be categorized into several stages, from early technological explorations to the current phase where AI generates content autonomously, indicating a significant shift in creative control [16][29] - The integration of AI in music production is transforming traditional workflows, allowing for faster and more efficient content creation, which is appealing to both creators and consumers [40][42] - Legal and ethical considerations surrounding AI-generated content are emerging as critical issues, particularly regarding copyright and the definition of artistic ownership in the AI era [42][44]

Core Insights - The report titled "2024 Chinese Digital Music White Paper" by Tencent Music Research Institute focuses on the overall landscape and trends of the Chinese digital music industry, showcasing industry dynamics and transformations through data insights, industry trends, cross-industry cases, and work reviews [1]. Group 1: Global and Chinese Market Data Insights - The global music market revenue is projected to grow from $70.7 billion in 2023 to $76.2 billion in 2024, with live performances remaining the largest revenue source at 46.1% [2][28]. - In China, recorded music revenue increased by 9.6% year-on-year, with online music paying users reaching 180 million, and Tencent Music's paying users rising to 121 million, reflecting a growing willingness to pay and improved service quality [2][34]. - The production of Chinese new songs surged to 1.351 million in 2024, driven by AI technology, while the number of long-tail songs (annual plays <1000) doubled, indicating intensified market competition and content diversification [2][45]. Group 2: Industry Core Trends - "County Narrative" has emerged as a focal point for creation, with local culture becoming a significant theme, as seen in works like "Factory" by Henan's rap artist, which resonates with societal issues [3]. - The "Silver Economy" is shifting from passive consumption to active participation, with the over-60 population exceeding 300 million, showcasing their potential in cultural consumption through events and user-generated content [4]. - The music festival market is experiencing intensified competition, with differentiation becoming crucial through unique artist lineups and cross-industry collaborations [5]. Group 3: Cross-Industry Cases - The game music for "Black Myth: Wukong" integrates diverse musical elements, showcasing the role of music in interactive storytelling and enhancing game promotion [6]. - Duolingo's music courses utilize gamification to lower barriers to learning instruments, achieving 3 million daily active users within six months [7]. - Innovative marketing strategies in live performances, such as the integration of nature and music festivals, demonstrate the potential of live music as a marketing medium [8]. Group 4: Annual Works and Industry Value Outlook - The white paper highlights the top ten artists, songs, and albums of the year, with notable performances from artists like Zhou Shen and Wang Sulong, indicating a balance between popularity and professionalism in music [9]. - The future challenges for the industry lie in embracing cross-industry thinking to break down barriers and adapt to technological changes while balancing content quality and commercial value [9].

Core Viewpoint - The audio industry is undergoing a phase of consolidation driven by the importance of scale in the content sector, with free business models demonstrating a strong ability to attract users and traffic, potentially breaking market size limits [1][4][25] Group 1: Company Developments - Tencent Music Entertainment (TME) plans to acquire Ximalaya for $2.4 billion, with the agreement potentially reached in the coming weeks [1] - TME's online music service revenue increased by 25.5% year-on-year, accounting for over 75% of total revenue, highlighting its growth strategy focused on online music [5][6] - TME's monthly active users for online music services have shown a decline from 620 million in 2022 to 570 million in 2024, indicating challenges in user retention [7] Group 2: Market Dynamics - The audio market has seen a lack of platform-level companies, with many players reverting to content production rather than platform development [1][4] - The audio industry faces a growth ceiling due to high content production costs and the uncertainty of monetization, leading to a trend where platforms must scale to survive [4][19] - The overall user base for online audio has stagnated at 335 million, with a notable decline in usage time from 58 minutes in 2020 to 25 minutes [22][24] Group 3: Competitive Landscape - Ximalaya, despite facing challenges in commercialization and high content costs, boasts a significant user base of 303 million monthly active users, capturing 60% of mobile listening time in China [11][18] - The competitive landscape is shifting, with platforms like Douyin and Kuaishou capturing market share from traditional audio live streaming services [8][20] - The trend of free models in content consumption has proven effective in other sectors, suggesting a potential for similar disruption in the audio market [25][26]

Core Viewpoint - The article discusses the launch of the MurekaO1 model and the MurekaV6 base model by Kunlun Wanwei, highlighting advancements in AI music generation technology and its implications for the industry [1][2]. Group 1: Product Launch and Features - Kunlun Wanwei released the MurekaO1 model, which is the first music reasoning large model globally, alongside the MurekaV6 base model that supports music generation in 10 languages [1][2]. - The MurekaV6 model incorporates self-developed ICL technology, enhancing sound field openness and improving vocal quality and mixing design [2]. - MurekaO1 is based on the MurekaV6 model and introduces the Chain-of-Thought (CoT) reasoning optimization, significantly improving music quality, creation efficiency, and flexibility, outperforming existing models [3][5]. Group 2: Technological Innovations - The MusiCoT innovation within MurekaO1 utilizes the CoT method to pre-generate overall music structure, enhancing coherence and precision in music generation compared to traditional autoregressive models [4][7]. - Mureka is among the first AI music generation platforms to offer open API services and model fine-tuning capabilities, marking a significant step in the industry [4]. Group 3: User Demographics and Business Model - The user base includes both C-end users who enjoy music and B-end professionals in film, gaming, and audio sectors, aiming to lower costs and improve efficiency [6][8]. - The business model offers free access for C-end users with paid options for enhanced features, while B-end services are charged through SaaS or PaaS [6]. Group 4: Market Potential and Strategy - The global digital music market is currently valued at approximately $40 billion, with expectations to expand to $100 billion as AI music generation technology matures [8]. - The company aims to establish a strong presence in the global market, leveraging technological advantages and international experience [8][12]. Group 5: Competitive Landscape and Differentiation - The company differentiates itself from competitors by focusing on B-end client needs and maintaining a high degree of globalization and multilingual support [9]. - The strategy emphasizes building a creator ecosystem and addressing misconceptions about AI-generated music within the industry [12]. Group 6: Future Directions - Future technological iterations will focus on providing comprehensive audio solutions and developing interactive music adjustment capabilities [11]. - The company plans to continue its commitment to open-source initiatives, enhancing product quality and establishing a strong reputation in the AI community [13].