DOCUMENTATION

From prompt to production-ready sound.

A practical guide to getting clear, useful results from the current beta.

Write the audible event

Name the source, exact action and perspective. “A heavy pump-action shotgun firing once from far away” is more useful than “epic gun sound.” AI Prompt Enhancement can translate and prepare the request without changing the event.

Choose the correct Sound Type

Sound Type changes prompt preparation and post-processing. Use UI for interface feedback, Weapons for firearms, Ambience for continuous environmental beds and Music / Tonal for instrumental material.

Shape the result

Style controls character, Duration defines the requested envelope, Intensity changes perceived force, and Seamless Loop is best reserved for ambience or repeating textures.

Compare before export

Choose 4, 8 or 12 variations for each generation. Preview them, rate useful directions and export lossless WAV or game-ready OGG.

Connect AudioMade

Your sounds, inside your AI workspace.

Connect with Codex CLI

Generate a key, then paste the command into your terminal. Requires Node.js/npm and Codex CLI. No installer file needed; npx fetches the MCP package automatically.

Terminal command
codex mcp add audiomade --env AUDIOMADE_API_KEY=YOUR_AUDIOMADE_API_KEY -- cmd /c npx -y audiomade-mcp

Restart your app and open a new conversation after setup.

The copied command contains your private key and may remain in shell history. Keep it private; never commit it to a repository. Generating sounds uses your credits. Existing connections stay active.