How to estimate subsetting savings before you commit
- Step 1Drop your font — Any TTF, OTF, WOFF, or WOFF2. The analyser parses the GlyphSet and projects savings.
- Step 2Read the per-subset projections — For each subset (latin, latin-ext, cyrillic, greek, vietnamese, symbols): glyph count in subset, projected WOFF2 size, savings percentage.
- Step 3Pick the right subset — Match your audience's languages to the smallest subset that covers them. Latin Basic for English-only is 80–95% savings; Latin Extended is 70–85%.
Frequently asked questions
How accurate are the projections?+
Within ±5% for typical Latin fonts. Variance increases for CJK and complex-script fonts where per-glyph compression is less linear. The actual subset run via Smart Font Subsetter gives an exact answer.
Should I always subset?+
If your audience's languages fit in a subset that's significantly smaller than the full font (≥30% reduction), yes. For fonts that are already minimal (icon fonts at 10 KB), the overhead may not justify the work.
What if I need multiple languages?+
Use unicode-range to ship multiple subsets, browser fetches only matching characters. Glyph Count projects each subset independently; combine subsets in the CSS via unicode-range. Each visitor downloads only what their content needs.
Privacy first
Every JAD Font tool runs entirely in your browser using opentype.js and the wawoff2 WASM Brotli encoder. Your fonts never leave your device — verified by zero outbound network requests during processing.