AMSAI Music Studio

AMS — AI Music Studio

AMS frequently asked questions

Updated: 2026-11-01

Is it free?

AMS uses no commercial AI inference API and requires no account. Internet access, electricity, and computing resources are your own.

Where is the music generated?

In the browser on your device. The server serves the application and does not receive generated audio. Only messages submitted through the contact form reach the support inbox.

Why do models need downloading?

Model weights are required for local inference. Large files download on first use; the browser may reuse them from cache later. Private mode, changing the origin, or clearing site data can require another download.

Does 10 seconds mean the waiting time?

No. Ten seconds is the audio duration. Processing can take several minutes and depends on your GPU, CPU, memory, and browser.

Can I use a track in commercial videos?

Models and components have MIT and Apache-2.0 licenses with their respective notices. Read the License page. These licenses do not guarantee the originality of an output or automatically prevent Content ID claims.

Is BPM exact?

Manual BPM is an instruction in the model description. The output tempo is not automatically measured or guaranteed to match.

How do I keep my tracks?

The library is local in IndexedDB. Use Download WAV to keep copies independent of the browser. Clearing site data can remove both the library and model cache.

Optional analytics

With your consent, Umami counts visits and usage events. We do not send prompts, audio, or seeds. You can use AMS without analytics.