Partly: in 2 runs, Claude Haiku 4.5's tailwindcss attribution reaches 4.0.0 (2025-01-21), inside tailwindcss 4, and stops there — tailwindcss was at 4.3.3 when this index last tested it (2026-09-28). That is ~1 months below the training cutoff the subject stated in those runs (2025-02). No finding is currently charged against Claude Haiku 4.5 on tailwindcss, but this is not a clean result: the 2 runs published here record 4 misses and 2 imprecisions that the index's charging rules do not turn into findings, each published in its run record. Read the boundary precisely: it is the newest release whose contents the model can correctly attribute to that release, not the newest tailwindcss feature it can use. Past it a model often writes working code with a newer API while naming the wrong release for it.
Answer class inside, decided by one rule applied to every page
of this kind: every measured boundary is at or above the first release of the major named in the question. Every figure below is read from
the dataset at build time; nothing on this page is written by hand.
2 measurements of this pair, all giving the same boundary. No two of them were drawn from an identical prompt, so the spread mixes prompt differences with draw variance.
Nothing is charged, and this is not a clean result: the runs below record 4 misses and 2 imprecisions that the index’s charging rules do not turn into findings. Each is published in its run.
The tailwindcss correction pack states what is true now for each corrected fact, with a primary-source citation, as markdown you can paste into a rules file (raw). What a correction pack measurably changed when one was tested — a pre-registered run on prisma and zod — is on the benchmark page, including where it changed nothing.
All of them at once: Which Claude model knows tailwindcss 4 best?
Other subjects on tailwindcss:
Claude Haiku 4.5 on the other libraries: