Fold Unicode accents to ASCII
Exacttext_unicode_foldDeterministic and authoritative — the canonical answer for this input.
ASCII-fold a string by decomposing unicode (NFKD) and dropping combining marks. 'café' → 'cafe'. Non-foldable characters (e.g. ß, emoji) are preserved as-is. Useful as a pre-clean step before matching or slugging. Processing multiple inputs? Use `text_unicode_fold_batch` to do up to 1000 in one call.
Parameters
textstringrequiredText to ASCII-fold.
Call it
curl -X POST https://api.snipget.ai/v1/text/unicode/fold \
-H "Authorization: Bearer sk_live_..." \
-H "Content-Type: application/json" \
-d '{"text":"Crème brûlée café"}'The Batch tab sends up to 1000 inputs in one call via /v1/text/unicode/fold/batch.
Example response
{
"input": "Crème brûlée café",
"folded": "Creme brulee cafe"
}FAQ
What does folding do?
It strips accents and diacritics from letters, turning 'Crème brûlée café' into 'Creme brulee cafe'. Useful before matching, searching, or building a slug, so accented and unaccented spellings line up.
How is this different from the slug tool?
Folding only removes accents and keeps the rest of the text, spacing, and case as-is. The slug tool also lowercases, replaces spaces and punctuation with dashes, and trims, to make a URL-safe string.
What about characters that have no plain ASCII form?
Letters like the German ß and symbols like emoji are kept as-is rather than dropped, so you do not lose meaning. It only simplifies characters that decompose cleanly to ASCII.
Is it reliable to run in a pipeline?
Yes. It is deterministic and authoritative: the same input always folds to the same output.