Record a few seconds of speech. Your voice becomes a unique, installable alphabet — the same voice always draws the same font.
Voice2Font analyzes a short recording of your voice — pitch, tempo, timbre, and a handful of other acoustic features — and maps those characteristics onto letterform parameters (weight, width, roundness, contrast, terminal style) to draw a genuinely unique, installable alphabet. The same voice always produces the same font, so it's a deterministic, personal signature rather than a random generator.
Making letters a little bolder or wider barely changes how a font feels. The big differences come from silhouette and texture — stroke endings, open or closed shapes, contrast — so that's what your voice controls most. A small, capped amount of hand-drawn wobble adds character without hurting legibility. Fonts stay upright; use your app's italic button if you want a slant.
Once generated, you can preview your font live by typing into the box below, then download it as a real, installable .ttf file. Your recording is processed in memory on the server and deleted immediately after the font is built — nothing is stored or reused.
English (Latin) letters. Korean (Hangul) isn't supported yet.
Download the .ttf file and open it. On Windows or macOS, click Install. It then appears in Word, design apps and anywhere else that lists your fonts.
The same recording always gives exactly the same font. A new recording of you is close, but not identical.