ElevenLabs adds video-matched music and sound effects with editable tracks
Instant Soundtracks reaches ElevenCreative, ElevenMusic and the API, with separate audio files for downstream mixing and support for using existing audio as a musical reference.
ElevenLabs released Instant Soundtracks on October 15, 2026, adding video-informed music and timestamped sound effects to its creator tools and API. Unlike a single finished mix, the output can be split into editable music and effects tracks, giving creators more control over revisions and downstream mixing. The API supports up to 10 videos per request, subject to combined limits of 200 MB and 10 minutes of footage. The release brings automated audiovisual scoring into ElevenLabs’ broader creative workflow, while leaving its timing and mood improvements as company claims.
01
The API’s `add_sfx`, `separate_tracks` and `use_audio_as_reference` parameters control effects generation, track delivery and musical guidance from existing audio.
02
Creators can adjust, move or replace effects independently in the Music Video Editor, or request a ZIP of separate music and effects files.
03
The model can use visual events and the uploaded video’s existing audio—such as its tempo or instrumentation—as separate sources of direction.
ElevenLabs has launched Instant Soundtracks, letting creators upload a video and generate music plus sound effects matched to the footage. Released October 15, 2026, across ElevenCreative, ElevenMusic and the API, the feature pairs automatic audio placement with separate effects layers or files that users can adjust rather than accept as one finished mix.
The footage supplies the cues
In ElevenCreative, the workflow begins with an uploaded video. The model composes original music, identifies moments that could call for a sound effect, then generates and places those effects at matching timestamps. ElevenLabs gives the example of a whoosh accompanying a logo at the end of an advertisement. The system chooses both the effect cue and its position on the timeline.
The company says the update improves timing for short-form content, including cuts, transitions and visible actions. It also claims more accurate changes in musical mood between scenes. Its example moves from electronic dance music or R&B during a fast-paced scene to acoustic guitar during a slower, more intimate one. Those timing and mood improvements are ElevenLabs’ characterization of the release.
Visual cues are not the only possible guide. Creators can use a video’s existing audio as a reference for the new score. The model can analyze genre, crescendos, tempo and instrumentation, then use those characteristics to shape the generated music. That gives the feature two sources of direction: what happens in the picture and the musical character already present in the uploaded video.
An automatic pass, then hands-on editing
The automatic placement does not make the result a single, fixed audio layer. In the ElevenCreative Music Video Editor, sound effects appear separately from the soundtrack. Users can review and adjust them independently, add more effects, move their timing or replace the music. They can also refine individual moments without starting the whole process again, then export directly from ElevenCreative when finished.
The API exposes three controls
For developers, the ElevenCreative Studio API exposes the release through three named parameters. They govern whether to generate effects, whether to separate the output tracks and whether to draw musical guidance from the original audio:
add_sfx enables the generated sound effects alongside the music, bringing the timestamp-matched effects workflow into the API.
separate_tracks=true returns a ZIP with music and effects as separate files, rather than requiring users to work with them together.
use_audio_as_reference enables the existing-audio reference option, carrying that source of musical direction into an API-based workflow.
The request limits apply to the uploaded material as a whole: the API supports up to 10 videos per request, with a combined maximum of 200 MB and 10 minutes of footage. The video count is therefore only one constraint; a request also needs to fit both the total file-size limit and the total duration limit.
Beyond the editor and API, ElevenLabs lists access through the ElevenMusic platform and Creative Engine, its route for adding music and effects to ads. These are access points for the same Instant Soundtracks release, not separate product launches.
Reader comments
Newest comments first. Replies stay oldest first.