As competition in large AI models gradually shifts from parameter races to real-world deployment capabilities, a number of Chinese AI startups focused on vertical scenarios are beginning to stand out. Among them is Ziyouliangji Information Technology, founded in 2023, which is attempting to redefine how ordinary people create music through AI.
Unlike many companies concentrating on general-purpose large models, Ziyouliangji has chosen to focus on the high-barrier music sector. Its core product, Hitto (YinChao), is an AI music creation platform powered by the company’s self-developed music foundation model. On Thursday, the company also showcased Hitto at the BEYOND Expo 2026, where it demonstrated its latest AI-powered music creation capabilities.
Users can generate complete songs simply by entering a sentence, uploading a photo, or describing an emotion. Through technology, the company hopes to lower the barriers to music creation and allow even people without musical training to compose songs easily.
Music democratization is a concept repeatedly said by Ziyouliangji. Jiang Tao, the company’s CTO, once shared that eight years ago he tried to create a song for his wife as a wedding anniversary gift, but the complicated and expensive traditional music production process left a deep impression on him.
That experience later became one of the key reasons he entered the AI music field. In 2024, as end-to-end music generation models became more mature, Jiang assembled a cross-disciplinary team combining algorithm expertise with musical backgrounds to begin developing a proprietary music foundation model.
Compared with text or image generation, AI music faces more complex technical challenges. Music generation requires handling ultra-long context, melodic structures, and emotional expression, while Chinese songs also involve unique linguistic features such as tones and soft pronunciations. Ziyouliangji believes this is one reason overseas AI music models have struggled to fully adapt to the Chinese-language market.
On the technical side, one of Hitto’s biggest strengths is its fully self-developed pipeline. The team adopted a hybrid AR+NAR architecture, enabling the model to maintain coherent song structures while also delivering refined local details. The model also has multimodal capabilities and can understand text, images, audio, and even video inputs in a unified representation space.
In the latest Hitto V3.0 release, the team further improved AI vocal performance quality. The model can now produce subtle singing techniques such as humming, vocal runs, and breathy vocals, while also adjusting emotional delivery according to lyrical content.
At the same time, the company focused on solving a common issue in AI music — songs that sound smooth but lack memorable hooks — by optimizing melody and arrangement generation to make songs more catchy and emotionally engaging.
Ziyouliangji’s current user base mainly consists of ordinary consumers. The platform has seen many life-oriented creative examples: truck drivers turning poems written on cigarette boxes into songs, families using photo-based song generation to document children’s growth, and users transforming heartbreak into musical expression. For many people, music is no longer viewed as a professional skill, but increasingly as a new form of emotional expression.
Beyond consumer entertainment, AI music is also entering commercial scenarios. The lyrics, composition, and vocals for AI For Good, the English theme song of the 2025 World Artificial Intelligence Conference, were partially generated using the Hitto model. The company has also begun collaborating with organizations in education, healthcare, and mental wellness to explore the use of AI music in areas such as therapeutic assistance and emotional support.
As AI music enters its next phase, industry competition is shifting from whether music can be generated to whether it can genuinely resonate with listeners. Ziyouliangji aims not only to build a music generation tool, but also to provide ordinary people with the ability to express themselves through music.
In an era where AI is becoming increasingly widespread, this emphasis on emotion and creative freedom represents one of the more distinctive directions emerging among Chinese AI startups.
