Queued
Waiting for compute
The job is accepted and waiting for the connected service to pick it up.
Workflow
How ANYANO works
Four stages, in order. Your music goes in, a personal Sound comes out, and generation becomes a way of exploring where your own material can go next.
Overview
The pipeline at a glance
UploadTrainGenerateExplore
Each stage has its own page in the studio and its own state, so you always know where a piece of material is: in your library, in a prepared dataset, inside a training run, or behind a finished Sound.
Step by step
The four stages in detail
- 01
Upload
Add songs, demos, stems, vocals, instrumentals, MIDI files and lyrics that you own or have sufficient rights to use.
Files land in your private library. ANYANO can analyse tempo and key, and separate a track into stems so you can train on isolated elements instead of full mixes.
- 02
Train
Group the material you want to learn from into a dataset, prepare it, then start a training run that produces a personal Sound.
You choose what the run should focus on and the basic hyperparameters. Before it starts you confirm you hold the rights to the material — including any voice recordings.
- 03
Generate
Select a finished Sound in the generator and describe what you want: prompt, lyrics, mood, structure, tempo and how far the model should stray.
Generations queue as jobs and appear in your generations feed with a waveform player when they complete.
- 04
Explore
Branch a generation into variations, extend a section you like, and collect what survives into a project.
This is where the work happens. A first generation is rarely the answer; it is a direction to follow or discard.
Status
What you see while a job runs
Training and generation are asynchronous. ANYANO reports the state the backend genuinely provides, and nothing more.
Processing
Preparing material
Files are being normalised, trimmed and converted into the representations training consumes.
Training
Building the Sound
The run is underway. Duration depends on dataset size and queue depth.
Ready
Usable
The Sound version is finished and selectable in the generator.
Failed
Stopped with a reason
Something went wrong; the reason is shown and your source material is unaffected.
Note
No fake precision
We avoid showing invented progress percentages when the backend cannot report meaningful progress.
Rights
Before you upload
You need to own the material or hold sufficient rights to use it for training — including rights in performances, samples and co-writes.
ANYANO asks you to confirm this before a training run starts, and asks specifically about voice recordings, because those carry additional considerations. This is deliberately a gate rather than fine print: the whole approach depends on artists training on their own material.
FAQ
Frequently asked questions
Related
Keep reading
Start with the music you already have.
Upload, train, and hear where your own catalog can go.