ElevenLabs releases SFX v3 for multi-sound scene prompts
The AMW Read
The release adds multi-sound prompting for an established audio AI company, with implications concentrated in sound production.
ElevenLabs releases SFX v3 for multi-sound scene prompts
ElevenLabs introduced SFX v3 on October 29 for prompts that combine several sounds, including music and vocals. The model is available in the Sound Effects tab of ElevenCreative, as well as Flows, Studio, Music Editor, the mobile app and ElevenAPI. The announcement does not specify whether users can edit separate stems, which export formats are supported or what access each plan provides.
The release expands ElevenLabs' sound-generation offering within a generative media market that spans several distinct production workflows. A single scene prompt gives creators a way to describe multiple cues and their intended sequence together. The result still needs editorial review: a completed generation does not establish whether each cue appears in the intended order or whether the transition works for the scene.
For builders evaluating SFX v3 through ElevenAPI or the creative tools, a useful test is a scene with an explicit sequence, such as a door closing before footsteps fade down a hall. Listen for both cues and their timing, then revise the relationship in the prompt if the sequence fails. Before choosing it for a production workflow, teams should also verify stem editing, export and plan access against their own requirements; those details are not established in the release described by the source.
