SFX v3

SFX v3 is ElevenLabs' newest text-to-sound-effects model, launched October 29, 2026. It targets creators who need multiple layered sounds from one prompt—crowd + ball + boos, or door + engine + footsteps—plus stronger music and vocal generation than SFX v2.

Launch: elevenlabs.io/blog/sound-effects-v3 · Docs: sound-effects capability page · API model_id: eleven_text_to_sound_v3

Key points

1. Named product: marketed as SFX v3 (API eleven_text_to_sound_v3); v2 remains default until you opt in.
2. Claims: more reliable multi-instruction prompts; cleaner high-frequency sounds; better music/vocals; available in Sound Effects tab, Flows, Studio, Music Editor, mobile app, and ElevenAPI.
3. Same-day traction: official @ElevenLabs post plus dozens of creator demos (pet-food ads, saloon ambience, ASMR) circulating on X in the 08:32–12:32 BJT window afterglow.
4. Access: limited-access in some regions (Unknown model if not entitled); commercial use needs a paid plan / credits.
5. Not Chimera: distinct from already-tracked ElevenLabs Chimera—this is the sound-effects line.

Why it matters

Searchable proper noun for a shipped audio model with official blog + API id the same day community clips went viral.

How to try it

1. Open ElevenCreative → Sound Effects / Flows and select SFX v3, or call ElevenAPI with model_id: "eleven_text_to_sound_v3".
2. Write multi-part prompts (ordered actions, soft vs loud layers) and compare against v2 on the same text.
3. Drop clips into Studio / Music Editor or your NLE; watch credit burn on specified durations.

Confirm regional access; keep v2 as fallback if the account lacks v3 entitlement.