Validate UTF-8 leaderboards

Top 20 submissions by each measure. Only submissions that passed the compliance suite are ranked. Each person can hold up to 3 places per board, or 5 on Fastest benchmark, so different techniques and languages can rank side by side; resubmitting the same Wasm counts once.

Showing only submissions that earned “Passes the UTF-8 validation compliance suite”. Select it again to show everyone.

Size against speed

Gzipped Wasm size against median benchmark time, coloured by the AI models each submitter declared.

616 B 977 B 2 ms 3.2 ms Gzipped Wasm size Median benchmark time patrickgwsmith · C · 776 B gzipped · 2.52 ms · gpt-6.1-sol
OpenAI: 1 of 1. 1 passing submission, each distinct Wasm once, on log scales. Lower left is smaller and faster; the dashed line joins those nothing else beats on both, labelled with language and AI model.

Times 1 of these in your browser, one at a time.

Speed, smallest first

Every benchmarked submission in order of gzipped Wasm size, to see whether getting smaller costs speed.

0 1 ms 2 ms 3 ms 45f455ae · patrickgwsmith · C · gpt-6.1-sol · 776 B gzipped · 18 memory pages · 2.52 ms 776 B 0 5 10 15 20 Memory (64 KiB pages) Median benchmark time
1 passing submission, smallest gzipped Wasm on the left, bar height the median benchmark time from zero. The 1 solid bars are each faster than everything smaller; muted bars have a smaller, faster submission to their left. Hollow dots are each Wasm's memory, in 64 KiB pages on the right-hand axis.

Times 1 of these in your browser, one at a time.

Do the machines agree?

  • Server
  • Server on the runner corpus
Time relative to the fastest submission (1×), on a log scale; further right is slower. Each machine's times are scaled for its overall speed so they line up. Hollow dots are visitors' browsers, which don't affect rankings.

By language

Fastest by language

Each language's fastest submission by median benchmark time, with its gzipped size and whether it uses SIMD.

  1. CSIMD 2.52 ms 776 B gzipped

Smallest download by language

Each language's smallest gzipped Wasm, which sets download time, with its median benchmark time. Open a language to see its next smallest submissions.

  1. C 776 B 2.52 ms

Fastest benchmark

Time to render the whole shuffled corpus once after warmup, timed in loops inside Wasm, with input copies subtracted. The server, browsers and other machines run the same harnessed Wasm the same way. Excludes setup and output validation. The corpus is generated (2889 generated inputs, 5.38 MiB in all, up to 255.5 KiB each): labels, prose, escaped HTML, URLs and dense references, with every size band counting for a similar share of the time. It is separate from the compliance suite and no input or output is larger than the suite's largest. The server ranks on its own corpus, which never leaves it; browsers run a public one and benchmark runners a third, all from the same generator with different seeds. They are rotated from time to time, so tune for the kind of input, not the cases.

  1. 1.
    patrickgwsmith
    Language
    C SIMD
    Compiler
    Homebrew clang 23.1.5
    AI
    gpt-6.1-sol
    2.52 ms · 2240 MB/s

Fastest suite

Total time to run the compliance fixture suite.

  1. 1.
    patrickgwsmith
    Language
    C SIMD
    Compiler
    Homebrew clang 23.1.5
    AI
    gpt-6.1-sol
    5710.81 ms

Smallest Wasm

  1. 1.
    patrickgwsmith
    Language
    C
    Compiler
    Homebrew clang 23.1.5
    AI
    gpt-6.1-sol
    1.3 KiB

Sign in to compare on your machine

Comparisons are saved with your GitHub account, so we can trust the results and set aside any that look wrong. We never ask for access to your repositories.