ElevenLabs has introduced Instant Soundtracks, a feature that scores an uploaded video with original music and timed sound effects in a single pass. Upload footage to ElevenCreative and the model writes a score to fit the picture, then watches for moments that call for an effect and places each one at the matching timestamp. The effects sit on separate layers from the music, so an editor can adjust them without starting over. The feature is also available in ElevenMusic and through the company’s Video to Music API, and its announcement was published on October 15 and updated on October 28.
What the tool actually does
The starting point is a finished cut, not a text prompt. Instant Soundtracks analyses the uploaded footage scene by scene, looking at pacing, movement and the moments where sound should land. It then generates an original score for the whole video, rather than dropping a stock track underneath it.
The aim is a soundtrack that follows the edit, instead of one the editor must cut around. A library track can sound fine on its own yet still feel disconnected from the picture, especially when scenes change quickly. Generating the music from the footage itself is meant to close that gap.
Short-form video is the clearest use case in the announcement. ElevenLabs says the update brings tighter timing around cuts, transitions and on-screen actions, along with more accurate mood changes across scenes. A fast-paced section can carry a driving track before the music shifts to something slower when the scene changes.
Separate layers, not a single mix
The most practical detail is how the result is handed back. In the ElevenCreative Music Video Editor, sound effects appear as separate layers from the soundtrack, so each effect can be reviewed and adjusted on its own. An editor can add more effects, change their timing, or swap the music without regenerating the whole soundtrack.
Once the mix is right, the finished video can be exported directly from ElevenCreative. That keeps the whole job inside one editor, rather than sending music and effects out to a separate mixing tool. For teams publishing daily, removing that round trip is where the time saving should come from.
A video’s existing audio can also be used as a reference for the new score. The model reads characteristics such as genre, crescendos, beats per minute and instrumentation from the original audio, then shapes fresh music with a similar character. That matters for brands that already have a sound and want a new track that still feels like theirs.
The move builds on ElevenLabs’ voice work, including its realtime speech-to-speech models, but applies the company’s audio expertise to picture rather than voice alone. Voice generation and video scoring are different jobs, yet both depend on timing sound precisely to something happening on screen or in conversation.
Where you can use it
Instant Soundtracks is available in three places: the Music Video Editor in ElevenCreative’s music tab, the ElevenMusic platform, and the Video to Music API. Creators can start in the editor, while developers can build the same capability into their own products through the API.
Developers get three new parameters to control the result. The add_sfx parameter adds matched sound effects, separate_tracks returns the music and effects as separate files in a ZIP archive for mixing elsewhere, and use_audio_as_reference shapes the score from the video’s own audio.
There are limits to note before planning a batch around it. A single API request accepts up to ten videos, with a combined total of up to 200 MB and ten minutes of video. Those are input limits for the request, not a promise about output length or sync quality on any particular clip.
The announcement does not publish pricing for the feature, so creators should check what their account includes before using the output in paid media. Licensing and plan limits matter here, because a track made for a personal test is not automatically cleared for advertising.
Why It Matters
Sound is often the last job on a short video, and the easiest one to rush. A mismatched track can make a polished cut feel unfinished, while hiring a composer or licensing music for every clip is slow and expensive for teams publishing daily. A tool that scores a cut and places effects in one pass could cut that bottleneck.
Editors should still judge the result rather than accepting it blindly. Timing, taste and brand fit are hard to automate, and a generated score can be in sync yet wrong for the story. The same need applies to AI-generated clips, where tools such as Google’s Flow and Veo video tools can produce silent footage that still needs a soundtrack.
For ElevenLabs, the launch also marks a shift in scope. The company built its name on realistic speech, but Instant Soundtracks moves it from generating individual audio assets to coordinating music and effects against moving picture. If the timing holds up on real edits, that is a useful step for short-form creators.
Background
ElevenLabs is best known as a voice AI company, and Instant Soundtracks extends that audio work into video post-production, where music and effects are usually handled as separate jobs.
Source: ElevenLabs announcement and DigitalToday report.
0 Comments