Ziyouliangji intends to utilize the AI music platform Hitto to enable everyone to become a song creator.

      As competition with large AI models gradually transitions from a focus on parameter counts to real-world deployment capabilities, several Chinese AI startups specializing in niche scenarios are beginning to shine. One such startup is Ziyouliangji Information Technology, established in 2023, which aims to transform how everyday individuals create music using AI.

      In contrast to many firms concentrating on multipurpose models, Ziyouliangji has decided to specialize in the high-barrier music industry. Its flagship product, Hitto (YinChao), is an AI music creation platform built on the company's proprietary music foundation model. On Thursday, the company showcased Hitto at the BEYOND Expo 2026, highlighting its newest AI-driven music creation features.

      Users can generate complete songs by simply entering a sentence, uploading a photo, or expressing an emotion. Through this technology, the company aspires to ease the process of music creation, allowing even those without musical training to compose songs effortlessly.

      Ziyouliangji frequently discusses the concept of music democratization. Jiang Tao, the company’s CTO, recounted that eight years ago, he attempted to write a song for his wife as a wedding anniversary gift, but was left with a lasting impression of the complicated and costly traditional music production process. This experience later sparked his interest in the AI music industry. In 2024, as end-to-end music generation models matured, Jiang formed a diverse team that combined algorithmic expertise with musical knowledge to develop an original music foundation model.

      AI music generation presents more intricate technical challenges compared to text or image generation. It necessitates managing ultra-long contexts, melodic structures, and emotional nuances, and crafting Chinese songs involves specific linguistic characteristics such as tones and delicate pronunciations. Ziyouliangji believes this is why foreign AI music models have faced challenges in adapting fully to the Chinese-language market.

      Technically, one of Hitto’s significant advantages lies in its entirely self-developed pipeline. The team implemented a hybrid AR+NAR architecture, enabling the model to maintain coherent song structures while providing detailed local precision. The model is also multimodal, capable of comprehending text, images, audio, and even video inputs in a unified representational space.

      In the latest release of Hitto V3.0, the team enhanced the quality of AI vocal performance. The model can now replicate subtle singing techniques such as humming, vocal runs, and breathy tones, while also adjusting emotional expression based on the lyrics. Additionally, the company has focused on addressing a common issue in AI music—producing songs that, while smooth, lack memorable hooks—by refining melody and arrangement generation to create more catchy and emotionally engaging tunes.

      Currently, Ziyouliangji's user base primarily includes everyday consumers. The platform has produced many creative examples from daily life: truck drivers converting poems written on cigarette packs into songs, families utilizing photo-based song creation to document their children's growth, and users channeling heartbreak into musical form. For many, music is increasingly seen not as a professional skill but as a fresh means of emotional expression.

      AI music is also making its way into commercial settings. Portions of the lyrics, composition, and vocals for AI For Good, the English theme song for the 2025 World Artificial Intelligence Conference, were generated using the Hitto model. The company has started collaborating with organizations in education, healthcare, and mental wellness to explore AI music's application in therapeutic assistance and emotional support.

      As AI music enters its next phase, the industry competition is shifting from whether music can be generated to whether it can truly resonate with audiences. Ziyouliangji aspires not only to create a music generation tool but also to empower ordinary people to express themselves through music. In a time when AI technology is becoming increasingly prevalent, this focus on emotion and creative freedom marks a unique direction emerging among Chinese AI startups.

      Jessie Wu is a technology reporter based in Shanghai. She covers consumer electronics, semiconductors, and the gaming industry for TechNode. You can reach her at jessie.wu@technode.com.