Firefly Generate Music for a video vs Sume music with inspect
Adobe Firefly can score an uploaded video. Sume music takes text and an optional image, so inspect the clip first, then write the brief and pass a still.

Adobe Firefly's Generate Music can take an uploaded video and score it from its pacing and tone. Sume's music request does not take a video: it takes a text prompt and an optional image_url. To score an existing clip on Sume, run video inspect on it, write the music brief from what you see, and pass an accepted still as the image.
Firefly details are from Adobe's community announcement; Sume details are from Music 1.0 and Video inspect.
What does Firefly do?
The post says Generate Music makes original background tracks from a text prompt describing mood, tempo, genre or instrumentation. You can also upload a video and Firefly analyzes pacing and tone to generate music to match. Generate Speech, for voiceover from a script, launched alongside it.
What does Sume take instead?
Music 1.0 accepts a text prompt and optional image conditioning, with no seed, temperature, guidance or duration parameter. The image must be a public HTTPS URL. Video inspect reads one media.sume.com clip already owned by the workspace; probe and stills are unbilled.
Inspect needs the clip on Sume first, so import it with POST /v1/media-imports if it lives elsewhere.
| Step | Firefly | Sume |
|---|---|---|
| Input | Upload the video | Import the clip, run POST /v1/video-inspect |
| Understand pacing | Firefly analyzes it | You read probe facts and stills |
| Write the brief | Automatic | You write tempo, key and arc in the prompt |
| Generate | Generate Music | POST /v1/music-router/generate with prompt and image_url |
How do I write the brief from an inspect?
Use the Music 1.0 brief axes: emotion, genre, tempo as a number, key, two to four instruments, an arc with one named moment, and era. Take the arc from the clip's cuts and the mood from a still. The docs say these are creative directions, not guaranteed settings, so listen to the result.
Close with "Instrumental, no vocals." unless there is narration. Check first that the clip has audio you intend to keep; see reference video silent-audio check before music.
Can I match the exact video length?
Not with a parameter: duration and duration_seconds are rejected. Steer length in the prompt, for example "a 30-second track" or a [0:00-0:30] marker, then verify the audio against the clip. Mixing the track under the video is a separate step covered in add background music to a video.
Sources
Related posts
More in Use cases
- AI Act deployer for an agency: the legal person, not each hand
Per the Commission FAQ, a company stays the deployer when contractors or freelancers operate the AI system for it. How that maps to one Sume workspace key.
- AI Act Article 50(1) and a rendered avatar video: the test
Article 50(1) needs a system built for genuine two-way exchange. A rendered, script-driven avatar video is one-way; check the four criteria. Not legal advice.
- AI Act closed-loop film previs: where preview ends, output begins
The EU FAQ puts closed-loop film production outside marking scope unless the output is final. Avatar preview stills are never captioned; the final MP4 can be.
- Making AI videos for clients: are you an AI Act deployer?
The Commission FAQ treats freelance or regular paid use of an AI system as professional, making you a deployer. Personal use is excluded. Not legal advice.
Written by Sume