🎙️ Quint — speech → semantic paragraphs
Speak (or paste text) and watch it transcribe and group into semantic paragraphs on a CPU. Code · pip install quintessentia
Click the mic, allow access, and talk. It buffers your audio, and when you pause it transcribes the sentence (Moonshine — a fast CPU model), restores punctuation, and folds it into the right semantic paragraph. English; first use downloads ~560 MB, so the very first response can take a minute.