Sonilo Brings Licensed AI Soundtrack Generation to fal.ai

The integration allows developers and creative teams to embed automated audio scoring into video platforms and editing tools. Instead of relying on text prompts, the model processes visual data to compose music that matches the exact duration of a clip. During internal testing, editors accepted the AI-generated tracks on the first attempt 87% of the time, with data suggesting these soundtracks can increase viewer engagement by 16%.

Sonilo delivers audio as a distinct track, ensuring that dialogue and sound effects remain untouched while music is applied. The system supports videos up to 600 seconds long and includes a text-to-music feature for users requiring specific stylistic control. To address common copyright hurdles in generative media, the company states its models are trained on licensed content, including the Shutterstock music catalog, providing a clear legal path for commercial use and monetization.

CEO Shawn Song noted that the partnership aims to move music production away from the traditional, fragmented process of post-facto licensing. This release follows Sonilo’s April integration with ComfyUI, signaling a broader strategy to position the company as a foundational infrastructure layer for AI-driven creative tools.