Sonilo video-to-music on fal.ai: 600 s of footage vs Sume's route

Sonilo's video-to-music model scores footage up to 600 seconds on fal.ai. Sume has no video-to-music call: inspect the clip, write a prompt, mix with Timeline.

5 min readSume
All posts

Sonilo announced on June 22, 2026 that its video-to-music model is live on fal.ai, and says it analyzes a video's pacing, motion and emotional arc and generates music matched to the video's exact duration, for footage up to 600 seconds. Sume has no video-to-music call. On Sume you describe the music yourself in a text prompt on the Music Router, optionally after probing the clip with video inspect, then place the track with a Timeline soundtrack. The Sonilo facts are from its press release.

What does the Sonilo release say?

It says each generation produces several soundtrack options for the same footage, the music is delivered as a separate audio track so volume can be changed without touching dialogue or effects, and original speech can be preserved. It also gives internal figures (an 87% first-attempt acceptance and a 16% engagement increase); those are the vendor's own, not independent measurements, and are not used here for any comparison.

Sonilo press release (June 22, 2026) vs Sume docs, read 2026-10-02.
QuestionSonilo releaseSume
InputVideo footage, or text on its text modelText prompt plus optional image_url
Length matchComposed to the video's exact durationNo duration field; steer length in the prompt, then trim in Timeline
Max footageUp to 600 seconds on fal.aiNot applicable
OutputSeparate audio trackAudio artifact, mixed under a spine with soundtrack
Options per runMultiple soundtrack optionsOne audio artifact per accepted job

How do I score a clip on Sume?

Probe the clip first. Video inspect reads one media.sume.com clip and returns probe facts, stills and optional transcription; it does not describe scenes semantically, so you read the stills and decide the mood. Write a brief with tempo, key, instruments and an arc moment, ask for a track about as long as the clip, then render with soundtrack (gain_db, fade_out_seconds, duck_db).

What about getting several options?

Submit more than one job with different briefs and an Idempotency-Key per intended job; each accepted job is a separate generation at the fixed Music price. Pick the one that fits the cut. This is a manual version of Sonilo's multiple options, and it is yours to run and review.

Does Sume say anything about licensing?

The Sonilo release says its models are trained on professionally licensed content and outputs are available for commercial use under its terms. The Sume docs pages used here do not make an equivalent statement, so check your own platform and rights requirements before publishing.

Sources

Related posts

More in Comparisons

All Comparisons posts

Written by Sume