Sonilo, a generative audio company founded by veterans of TikTok's AI and music teams, announced on October 14, 2026, that it has raised US$11 million in funding. The round was led by B Capital with participation from Redpoint. The company is building proprietary video-native music and sound models that adapt to the pacing, emotion, and story of each video.
Video-Native Sound Generation
At the core of Sonilo's platform is its Sound World Model, a proprietary multimodal AI system that connects video understanding, audio generation, and temporal alignment. Instead of treating music and sound effects as isolated assets, the model reads the timing, movement, and emotional arc of a scene. Creators can generate a synchronized, commercially ready soundtrack by dropping in a video, describing what they want in text, or combining both.
Funding Purpose and Strategic Backing
The new capital will be used for compute infrastructure to scale model performance, go-to-market expansion across creator and developer platforms, and growth of its music licensing operations. B Capital General Partner Daisy Cai said generative audio will become an essential part of the AI-native video stack. She noted that Sonilo's multimodal technology gives creators two ways to generate music and sound effects, either through text descriptions or by providing video that the model reads directly.
Founding Team and Market Focus
Sonilo's founding team brings together expertise in multimodal understanding, audio generation, and music licensing. CEO Shawn Song is a CMU-trained researcher and former multimodal AI tech lead at TikTok, while CTO Alex Yin holds a PhD in Computer Music from the University of York. COO Keli Li built and led global music at TikTok, and CMO Trista Taylor has helped incubate consumer AI apps featured in a16z's Top 50 rankings.
Creator Workflow and Product Offerings
Sonilo's platform generates music and sound effects directly from footage or text for short-form and short-drama creators, developer platforms building AI video workflows, and enterprise production teams. The platform also offers AI dubbing with synchronized lip-syncing, enabling short-drama and content studios to localize video across languages without manual post-production. Early distribution partners include TapNow, an AI creator aggregator, and ComfyUI.
Licensing and Artist Compensation
Sonilo is building its commercial music infrastructure around licensed source material, with Shutterstock as an early licensing partner. Its models are trained on licensed music datasets, with licensing fees and revenue share flowing to rights holders and artists. The company is finalizing master recording and publishing deals with major rights organizations, with agreements expected to close in the coming weeks.
Leadership Perspective
CEO Shawn Song highlighted the limitations of text-first audio-generation tools, noting that creators often spend hours describing music, aligning it to video moments, and finding sound effects. He said Sonilo was built to understand the visual story and generate music and sound effects that naturally fit the scene. Song added that the future of video involves more than generating better pictures, but also making those worlds feel real through audio that fits perfectly.
Consumer Launch and Industry Outlook
Sonilo plans to launch a consumer app in Q4 2026, bringing its video-to-music generation technology directly to individual creators. The company is headquartered in San Francisco and backed by B Capital. Its model development is built around the creator's actual workflow rather than industry benchmarks, a philosophy the team developed while building recommendation and generation systems at platform scale.
The funding positions Sonilo to advance generative audio as a core layer of the AI-native video stack. With a founding team from TikTok's AI and music divisions, early licensing partnerships, and a workflow built around video, the company aims to make synchronized soundtracks accessible to creators and platforms. By tying audio generation to the visual story rather than isolated prompts, Sonilo is pursuing a distinct category in generative media.