
Sound for a video scene
Write the main action and the sounds around it, and get a layered track that fits a generated or filmed clip.
ElevenLabs' newest text-to-sound-effect model: whole scenes from one prompt, cleaner audio, up to 30 seconds.
باقات رصيد أو خطة شهرية، برصيد واحد لكل النماذج على ZOOOP. · اشحن عند الحاجة؛ يبقى رصيد الباقات حتى تستخدمه.
Powered by ElevenLabs's API on ZOOOP
Describe several sounds at once, such as a crowd cheering, a ball dropping and the same crowd booing, and more of them make it into the result than with V2.
Generate anything from a 1-second hit to a 30-second ambient bed in a single pass, longer than the 22 seconds V2 allowed.
Effects that pair a musical phrase or a vocal performance with everyday sounds, like a street musician over traffic or a chant in a stadium, come out more convincingly.
Gongs keep their resonance, chimes keep their ring and whistles hold their tone without turning harsh or crackly.

Write the main action and the sounds around it, and get a layered track that fits a generated or filmed clip.

Impacts, whooshes, UI clicks, pickups and stingers, generated on demand instead of searched for in a library.

Rain on a window, a busy cafe or a forest at night, up to 30 seconds per clip for background layers.

Footsteps on gravel, cloth rustle, a glass set on a table: describe the foley you need and generate it.
Pick the right audio model. Your credits work everywhere on ZOOOP.
Open Sound Effects V3 from this page or pick it in the Audio tools.
Describe the sounds you want in English, starting with the main action, then what surrounds it.
Set the duration, from 1 to 30 seconds.
Generate, then adjust the description and generate again until it sounds right.
Sound Effects V3 is the model to reach for when you need sound for a scene and don't want to assemble it from a library. Describe what you want to hear and it generates the audio: one-shot effects like impacts and UI clicks, foley like footsteps and cloth rustle, and ambient beds like rain and room tone. ElevenLabs released it on October 29, 2026 as the successor to SFX v2.
The flagship change is multi-sound prompts. Big scenes rarely come down to one sound: a football clip needs a crowd cheering, a ball dropping and the same crowd booing seconds later. V2 tended to pick one or two of those; V3 follows several instructions within a single prompt, so more of what you describe ends up in the clip. The practical workflow is to write the main action first, add the surrounding sounds, listen, then adjust the description.
The second is effects with music and vocals. A street musician playing over traffic, a chant echoing around a stadium, a jingle under a door chime: prompts that mix a musical phrase or a voice with everyday sounds come out more convincingly than before. The third is plain sound quality: ElevenLabs reports fewer harsh or crackly artifacts on sounds like gongs, chimes and whistles, so each sound keeps its character whether it sits in front of the mix or behind it.
On ZOOOP a clip runs 1 to 30 seconds, up from 22 seconds on V2, and the prompt can be up to 450 characters. Write prompts in English for the best results.
Where it's weaker: seamless looping is not available on V3 yet; ElevenLabs documents it for V2 only, so for a bed that has to repeat without a seam, use Sound Effects V2 while it lasts. V3 does not write full songs; use Lyria 3 Pro for music and Minimax Music V2.6 for songs with lyrics.
A reasonable mental model: make Sound Effects V3 the default for any effect, foley or ambience, especially when a scene needs several sounds at once. Reach for Eleven V4 for narration and Lyria 3 Pro for music.
ElevenLabs released SFX v3 on October 29, 2026. It follows prompts that describe several sounds more reliably, handles effects that include music and vocals better, and produces cleaner audio with fewer artifacts than V2. On ZOOOP it also allows clips of up to 30 seconds instead of 22.
From 1 to 30 seconds in a single generation. The prompt can be up to 450 characters.
Not on ZOOOP yet. ElevenLabs currently documents looping for V2 only, so the loop switch is held back on V3 until it is confirmed to work.
Yes. It is priced per second at the same low rate as V2, so you can generate every cue in a project instead of licensing a stock pack.
Sound Effects V3 makes effects, foley and ambience. For full music tracks use Lyria 3 Pro; for spoken voiceover use Eleven V4.
Sound Effects V3
Prompt*
المدة*
ستظهر أجيالك هنا بمجرد تسجيل الدخول.
إنشاء حساب مجاني.