same! but if you get it inevitably wrong the first time it gives you the pinyin. but i struggled to get it to transcribe the consonants I was making let alone the tones. i'm pretty sure i'm not as bad as that!
It looks like it requires a MIDI which it then converts to sound like it’s coming out of a Gameboy.
Here’s what the FAQ says:
> How it works
>
> Search a song, pick a MIDI source, hit Generate. The Wario Synthesis Engine analyses the MIDI and resynthesises it using Web Audio oscillators tuned to mimic the Game Boy's 4-channel sound chip. All processing runs in your browser.
I’ve been using Gemini’s generous free tier in Gemini CLI. It’s nice to have other free options, even if it’s for a limited time (and with the caveat that the data will be used to improve the models).