Hi HN! I’m a part-time music producer who got tired of rummaging through sample packs to find just the right sound. I decided to build a plugin that generates audio from text prompts using a diffusion-based model. You can try it in your DAW (VST3/AU) or via a web app. The plugin itself is free to install, but you must pay for any generated audio (credit based at the moment, no subscription required).

*Why I Built This* After messing around with generative AI for images, I kept wishing something similar existed for music production. A weekend hack turned into a journey of building a real plugin, dealing with DAW quirks, GPU infrastructure, and figuring out a fair monetization model.

*What Makes It Different* - Integrated directly into DAWs: you can drop generated clips right onto your tracks. This allows you to stay in flow state - Quick iteration: type a prompt, wait a few seconds, and get your WAV. - Credits-based: $0.01 per second of generated audio.

*Ways to Try It* 1. Download the beta plugin on macOS (VST3/AU). Windows support is in progress. 2. Or use the web app (no installation, but you can’t drag directly into a DAW).

[*Website*](#) (https://text-to-sample.com)

I’d love to hear your feedback—especially if you produce music or have ideas for how to make AI-generated audio more useful. Feel free to ask questions; I’m around to discuss details of the implementation or to chat about generative AI for music in general. Thanks!