Sonilo’s video-to-music model is now live on fal.ai, giving creators a way to quickly generate original, commercially licensed soundtracks from video footage
SAN FRANCISCO, June 22, 2026 /PRNewswire/ — Sonilo today announced that its video-to-music model is now available on fal.ai, enabling users to generate a custom, licensed soundtrack for video in seconds.
Related posts
fal.ai * SONILO Partnering
The launch makes Sonilo available through fal.ai’s generative media infrastructure, expanding access to developers and creative teams building AI video products, editing tools, creator platforms and multimodal workflows.
Sonilo automates what has traditionally been one of video production’s most time-consuming steps: finding, syncing and licensing music for commercial use. Instead of requiring users to search stock libraries or write text prompts, Sonilo analyzes a video’s pacing, motion and emotional arc, then generates matching original music automatically. Each soundtrack is composed to match the video’s exact duration, helping creators avoid manual editing, trimming, or looping.
Each generation produces multiple soundtrack options for the same footage, giving creators a range of emotional directions to choose from instantly. In internal tests, editors accepted the model’s first attempt on 87% of clips. Videos scored with Sonilo showed a 16% increase in engagement, meaning audiences watched longer, interacted more and were more likely to share.
The generated music is delivered as a separate audio track, allowing editors to adjust its volume independently without affecting existing dialogue, voiceover or sound effects. Sonilo can also preserve original speech from source footage, making it easier to add music without compromising narration, interviews or spoken content. Through fal.ai, the model can score videos up to 600 seconds long. Sonilo’s text-to-music model is also available on the platform for those who prefer to work from a prompt, including segment-level controls that let creators define different musical styles, moods and structures across sections of a track.
“AI video creation is moving faster than ever, but music is still often treated as something creators have to search for, edit around and license after the fact,” said Shawn Song, CEO of Sonilo. “By launching on fal.ai, we are making it easier for developers and creative platforms to bring prompt-free, licensed soundtrack generation directly into their video workflows.”
Every Sonilo model is trained on professionally licensed content, including Shutterstock’s music catalog, with musicians compensated for their participation. Outputs generated by Sonilo are available for commercial use under the applicable deployment terms, giving video makers a clearer rights foundation for publishing and monetizing content.
The fal.ai launch follows Sonilo’s integration with ComfyUI in April, as part of a broader effort to establish the company as the go-to music infrastructure layer for the next generation of AI creative tools, with further platform integrations planned.
Both models are available now on fal.ai, with API access for developers building at scale.
- Video-to-Music: fal.ai/models/sonilo/v1.1/video-to-music
- Text-to-Music: fal.ai/models/sonilo/v1.1/text-to-music
About Sonilo
Sonilo is the world’s first professionally licensed video-to-music AI platform. Powered by its proprietary v1.1 model, it composes original music directly from footage or text in seconds. Sonilo is headquartered in San Francisco and backed by B Capital.
Website | X | LinkedIn | Instagram
fal.ai is a generative media platform providing high-performance inference APIs across image, video, audio and 3D models. It serves more than 2.5 million developers and enterprises in industries ranging from gaming and e-commerce to creative production.
