Skip to content

ElevenLabs Releases SFX v3 Sound Effects Model

ElevenLabs

The model generates multi-stage audio scenes, music, and vocals from single text prompts with improved prompt adherence over SFX v2.

Key points

  • Follows prompts describing multiple sequential or layered audio events step-by-step.
  • Reduces audio artifacts on sustained or resonant sounds such as gongs, chimes, and whistles.
  • Synthesizes music and vocal elements paired directly with environmental sounds.
  • Available through ElevenAPI, the mobile app, and ElevenCreative tools including Studio, Flows, the sound effects tab, and the music editor.

Why it matters

Audio creators and developers can generate complete, sequential sound scenes from single text descriptions without stitching separate audio clips.

Sources