Layer supports audio generation powered by ElevenLabs, enabling you to create both voiceovers and sound effects directly in your creative workflow.
Generate audio with the Agent
In any chat, describe the audio you want — "a cheerful coin-collect sound effect," "a calm narrator voice reading this line" — and the Agent generates it directly.
Generate audio with the direct form
From inside a session, click Generate audio on the bottom toolbar to open the generation form.
Voice generation
Voice generation is powered by ElevenLabs' multilingual model, known for its emotional range and lifelike speech.
Select Voice as your audio type.
Enter your script — to generate in a different language, simply type your script in that language.
Choose a character voice.
Adjust advanced settings if needed — options like stability, speed, similarity, and style exaggeration are typical for ElevenLabs-powered voice generation, letting you trade off consistency against expressiveness.
Click Generate to produce your voiceover.
Sound effects generation
Select Sound Effect as your audio type.
Write a description of the sound you want (e.g. "glass breaking," "sci-fi door opening," or "underwater bubbles").
Set the duration — Auto, or a custom length in seconds.
Click Generate to create your sound effect.
Want to edit a segment of the audio? Use the inpaint tool within the lightbox by double clicking the audio asset.
Generate audio inside a Workflow
For repeatable, multi-step pipelines, audio generation is also available as nodes on the Workflows canvas — Generate Audio, plus Change Voice, Isolate, Remove Noise, and Separate Audio for cleaning up or transforming existing audio. This is the right choice when audio is one step in a larger, automated pipeline rather than a one-off generation. See How to Use Workflows in Layer for the basics of building a workflow.
What's next
Drive generation directly → Generate and edit assets directly
Automate a repeatable pipeline → How to Use Workflows in Layer