Audio Editing

Trim audio locally with the Wonda CLI and route AI audio cleanup through direct audio endpoints.

Audio editing is CLI-first. wonda edit audio runs local ffmpeg operations, uploads the result, and prints the output mediaId.

The retired POST /api/v1/audio/edit endpoint returns 410 Gone with a CLI migration hint. Move audio trim scripts to the CLI command below.

Setup

npm i -g @degausai/wonda
wonda auth login
wonda doctor

Trim audio

Use audioTrim with --audio-media. The value can be a Wonda media ID or a local file path.

wonda edit audio --operation audioTrim \
  --audio-media 019e1b73-32a4-8c9e-8bbf-3661a14a5716 \
  --params '{"trimStartMs":1000,"trimEndMs":8000}' \
  --wait -o ./trimmed.mp3

Parameters:

Parameter	Type	Description
`trimStartMs`	int	Start of the kept range, in milliseconds.
`trimEndMs`	int	End of the kept range, in milliseconds. `0` means until the end.

Without --output, the CLI prints:

{
  "mediaId": "019e1b73-32a4-8c9e-8bbf-3661a14a5716",
  "status": "succeeded"
}

Audio AI operations

Audio enhancement and voice extraction are direct inference workflows, not editor operations:

Need	API endpoint	CLI command
Denoise or enhance audio	`POST /audio/enhance`	`wonda audio enhance <video-or-audio>`
Isolate vocals from a track	`POST /audio/extract-voice`	`wonda audio extract-voice <video-or-audio>`

Both API endpoints expect an audio media input. The CLI commands can accept a video or audio input and handle local audio extraction before calling the API.