Split text into parts

Split by grapheme count, a literal delimiter or a page marker, then download the complete text and individual parts.

Each input: 128 KiB, 65,536 UTF-16 units and 5,000 lines; 256 KiB combined. Full output plus parts: at most 512 KiB, with 5,000 lines in the full text and each part, and 100 parts; 5-second deadline. Output converts CR/CRLF into LF. The original remains in the field.

This selection controls comparison, letter case and cluster boundaries; it does not change the site interface language.

Count mode uses 1–65,536 clusters and preserves graphemes. A literal delimiter can split a cluster when it matches inside it; the delimiter or U+000C is removed only from parts, and empty parts remain. Complete text retains delimiters. Page boundaries are not inferred from screen layout.

How to use

Paste text and select its processing language and splitting mode. Enter a cluster count or delimiter when needed. Process and choose a part to preview or download; full download and copy retain the complete text.

Try an example

Splitting A|B||C by | produces A, B, an empty part and C; the full-text download remains A|B||C.

Support & limits

Count mode measures grapheme clusters and does not cut joined emoji. A literal delimiter can match inside a grapheme and split it; delimiters are removed only from parts, and empty parts are retained. Page mode splits only at U+000C, not at visually displayed pages. CR/CRLF become LF and complete output retains splitting delimiters. More than 100 parts or excess output budget is rejected without truncation.

Common questions

Can joining the parts always recover the text?

Count-based parts can be joined directly. Delimiter or page splitting requires reinserting the delimiter between parts; full copy and download already retain it.

Related tools