Benchmarks

Formula support table

RaTeX (Rust + WASM) vs KaTeX 0.16.45, row-by-row for the main golden suite (embedded page data; sourced primarily from `tests/golden/test_cases.txt` and related lists—line count follows the repo; includes mhchem \ce / \pu rows where present). Offline cells use pre-computed ink IoU vs KaTeX reference PNGs; the RaTeX column is computed live in your browser from the loaded WASM. bussproofs `prooftree` is tracked separately in the Proof trees gallery because KaTeX has no `prooftree` renderer.

Data source

Batch offline scores and aggregate counts are regenerated in CI runs and may lag the latest main by a few commits. The per-row RaTeX value always reflects the WASM you just loaded. For a single-formula sanity check, use live comparison.

—

score ≥ 0.9

—

0.8–0.9

—

0.5–0.8

—

0.3–0.5

—

< 0.3 or error

—

avg score

Initializing…

≥ 0.9 0.8–0.9 0.6–0.8 0.4–0.6 0.3–0.4 < 0.3 / error Offline IoU vs KaTeX PNGs

Initializing…

Scores come from offline golden comparison (RaTeX server PNG vs KaTeX reference). The RaTeX column is rendered on demand with WASM on this page; fonts match the gallery setup. Again: golden pipeline outputs may lag the repo—see the note above.

#	LaTeX source	KaTeX (reference)	RaTeX (WASM)	Score