DEV Community

IA Compa
IA Compa

Posted on Originally published at iacompa.com

Mirelo Sound Agent: Mirelo AI's AI sound design agent edits a video's sound from plain-language prompts

Mirelo AI has launched Sound Agent, a new way to sound-design a video by describing the sound you want in plain language.

Instead of working track by track in a timeline, you type a prompt such as "Sound design this clip." The agent then splits a video's audio into speech, sound effects and music stems, and edits them as separate layers.

What changed

Sound Agent is designed for the sound decisions that usually take the longest in post-production. In one prompt, it can:

  • Generate sound effects for a video
  • Remove unwanted music while keeping the sound effects
  • Remove a soundtrack entirely
  • Dub dialogue into another language
  • Split speech by speaker
  • Add a specific sound at a specific time, for example "At 2.4 s, add a heavy thud."
  • Change a score's mood over time, for example "Start calm, then turn intense."

The agent also works from reference audio. You can upload a sound, hum a melody or describe the sound you want, and it will find or generate a matching option.

Metrics

Metric Sound Agent
Sound effects generation From prompt, reference audio, humming or description
Dialogue dubbing English, Spanish, French, Mandarin
Price (Creator) $20 / month
Price (Studio) $60 / month
Price (Enterprise) Custom
Price (per video) $1.50
Metric Value
Credit price $1.50
Credits per video sound-design 1
Cost per minute (30 fps) $0.75
Cost per minute (24 fps) $0.60
Cost per minute (60 fps) $1.50

Why it matters

Sound Agent turns sound design from a manual, track-by-track job into a prompt-driven one. For creators and editors, that could shorten one of the slowest parts of post-production — especially for short-form video, where sound often decides whether a clip lands.

Sources

Originally published at iacompa.com

Top comments (0)