Every kind of sound a video needs can be generated in place: music with structure, voiceovers in natural voices, effects, and foley matched to a clip.
Pick the sound type On the Generate tab, switch the output type to Audio. Model cards cover music (with optional lyrics), text-to-speech voiceover, sound effects, and video-matched foley.
Describe it Music takes a style prompt (and a separate lyrics field with verse/chorus tags). TTS takes your script and a voice choice. SFX takes a short description and duration.
Match foley to a clip The foley model reads a video clip's source window and generates sound that fits the motion: footsteps, ambience, impacts.
Drop it in the mix Generated audio lands in your library; drag it onto a lane and duck it under any voice like normal footage.
Tips
Costs are shown per run before you generate; longer durations cost proportionally more.
Generate two or three music variants and keep the best: audio runs cost very little.
The voice model reads your script as written, so spell out numbers and acronyms the way you want them spoken.