Sonilo launches its video-to-music model on fal.ai, enabling rapid generation of licensed, original soundtracks tailored to video duration and emotion, and expands access for developers and creative teams building AI video products and multimodal workflows.

Internal tests show editors accepted Sonilo’s first attempt on 87% of clips, with videos featuring Sonilo music achieving a 16% lift in audience engagement.

Sonilo is a San Francisco–based company backed by B Capital, operating a licensed video-to-music platform, with fal.ai serving more than 2.5 million developers and enterprises.

The release is framed as part of a broader move toward multimodal AI that can understand and generate content across formats, with the AI-generated music market projected to approach $4 billion by 2033.

All Sonilo models are trained on professionally licensed content, including Shutterstock, and outputs are commercially usable under deployment terms with musicians compensated.

The collaboration positions Sonilo within fal.ai’s ecosystem and sets the stage for further platform integrations.

Fal.ai is a generative media platform that powers these integrations for developers and enterprises.

The two key models available are video-to-music at fal.ai/models/sonilo/v1.1/video-to-music and text-to-music at fal.ai/models/sonilo/v1.1/text-to-music.