Wavemaker’s composer has three deliverables: Video, Images, and Audio. Audio mode produces downloadable files you can play and download in chat — not a video soundtrack, and not a Remotion composition.
Not the same as video sound. On-screen dialogue, narration, and music inside a video are covered in Audio: voice, dialogue & music. Use Audio mode when the deliverable itself is an MP3 (radio spot, jingle, VO file, podcast episode).
Unbound Send routes Audio briefs onto catalog audio-* spines (Decide-for-me). You can also Try an audio workflow from Discover.
Purposes
| Purpose | What you get | Typical use |
|---|---|---|
| Music | Song, jingle, or instrumental bed | Beds, hooks, short jingles |
| Voice | Single-speaker narration | Scripts, explainers, announcements |
| Spot | Radio / podcast ad with duration packs | :06 / :15 / :30 / :60 traffic |
| Podcast | Topic episode | Shows, interviews, branded episodes |
Paste a brand site URL on Spot briefs when you want the name and CTA grounded in real pages.
In chat
- Describe an audio brief (or pick Audio if the composer shows output modes) and Send.
- When the set finishes, play / download from OUTPUTS or Files.
- Iterate (regenerate, change style/voice) without opening a video pipeline.
Credits
Audio uses the same reserve → settle model as video. Cost scales with purpose, length, and variants. Failed infra settles at zero; cancel settles to actual. See Credits & billing.
API & agents
POST $PLATFORM_API_URL/runs
# bind an audio-* catalog workflowId + inputsRetired: POST /api/v1/audio (HTTP 410). See API & MCP.