Speaking MIDI

Turn speech or music into piano notes. Listen, follow the piano roll, and download the MIDI. This is experimental sound approximation: words may become difficult to understand.

Start with

Audio → MIDI Text → speech → MIDI

Input audio

Fitting style

Speech keeps dense short notes. Instrumental uses fewer, longer notes.

Speech (dense) Instrumental (sparse)

CPU inference is supported. Speech generation can take several minutes; piano fitting runs on CPU. Try the ready-made examples for instant playback. Clips are processed up to the selected duration.

Choose an example or create a new fit.

1 · Source used for fitting

Download MIDI

Poetry voices · select to listen, then generate piano MIDI

Selected text

How it works: render piano templates → fit dense short notes or sparse held notes → convert strengths to MIDI velocities → render the MIDI with the same piano sound. Algorithm and experiments · Qwen3-TTS