Video Sound — guide skill for voice & voiceover

For AI agents: find the right skill in one call — curl -s "https://varg.ai/how-to/api/search?q=<what the person wants>" returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://varg.ai/how-to/llms.txt. Every page has a markdown twin: add .md to its path. Before generating anything, show the person the options on the how-to page. From a terminal: curl -fsSL https://varg.ai/install | sh installs the vargai CLI (https://varg.ai/how-to/cli.md): vargai search and vargai install <slug> with no key, every varg model after vargai login.

Sound for any video: real AI voices picked from candidates, lines with pauses and breaths, effects on the frame, music cut to bars, a mix checked by script.

What it does

An agent can't hear, so this skill makes the sound of a video with a plan, people's ears where taste is needed, and numbers everywhere else. It is built from 480 sourced findings (Walter Murch, Randy Thom, Ben Burtt, Richard King, trailer editors at AV Squad, NPR and CBC radio craft, voice directors, the ElevenLabs, Google, Kling, Seedance and OpenAI docs, HyperFrames, earshot, Cinetic) and paid tests on varg. The routine: a sound plan before any picture; lines written for the ear with pauses, breaths and small actions in each TTS model's documented syntax; 3-5 voice candidates rendered, measured and put on one page a person picks from; music prompted with genre, mood, instruments, BPM and era, measured, fitted to length on downbeats, with stops and drops made in the edit; effects only where something happens on screen (pro launch films leave about 85% of cuts without one); room tone under everything; then scripts/sound.py builds the mix from a cue sheet, ducks the music hardest in the voice band, masters to the platform and checks loudness, true peak, voice over music, effects within a frame, cuts on the beat and dead air. Tests behind it: bare text on eleven_v4 ran 231 words a minute with a 0.36 s pause, directed text 194 and 0.92 s with no tags read aloud; ElevenLabs Music held 120 BPM 3 of 3 times but made the planned silence 0 of 3; a designed launch soundtrack passed 8 of 8 checks where the plain one had the music 6.6 dB over the voice. Profiles for launch films, cinematic AI films, music videos, trailers and ads, UGC and talking characters, narrated explainers and code demos; the whole skill also comes as one copy-paste prompt.

Use when

Not for

Install

Example output

https://s3.varg.ai/files/194b94d0-58ff-4e8b-91de-4f1959d04013/file_32cbafbd0791.mp4 — The 20 s launch soundtrack from the skill's example in varg.ai colours: voice lines, effects and cuts over the waveform and spectrogram (2026-11-03) (credit: varg, https://github.com/vargHQ/skills/tree/main/video-sound/examples/launch-20s)

More voice & voiceover skills