Split PDF.拆分 PDF。

Split PDF on StayFile opens one PDF in this tab and copies the pages you ask for into a new download. It is a free page from Haimc (海明创), with no account and no upload. Two choices sit under the file. “Pages I list” writes the pages you type, in the order you type them, into one PDF named stayfile-pages.pdf. “One PDF per page” writes each page into its own PDF and packs those files into stayfile-pages.zip. The original on disk is not modified.

StayFile 的拆分 PDF 在这个标签页里打开一份 PDF，把你要的页面复制进一份新的下载。这是海明创（Haimc）的免费页面，没有账号，也不会上传。文件下面有两种做法。“按我写的页码”把你输入的页面，按输入顺序，写进一个名叫 stayfile-pages.pdf 的 PDF。“每一页一个 PDF”则把每一页写成单独的 PDF，再装进 stayfile-pages.zip。磁盘上的原件不会被修改。

Page numbers start at 1, in the order a normal viewer uses. The number printed in a footer can disagree with that index. If you are looking at the third page and the footer says “iv” or “12 of 40”, type 3. The box is filled with a short starter, such as 1-3 on a longer file, or 1 when the file has a single page. Replace that starter with the pages you actually want. Choosing another PDF replaces the one already open. If you drop several files, only the first is opened. This tool does not split a folder in one pass.

页码从 1 起，按普通阅读器的顺序。页脚上印的数字可能和这个序号不一致。如果你看着的是第三页，页脚却写着“iv”或“12 / 40”，请输入 3。输入框里会先放一个很短的起步范围，比如较长的文件先写成 1-3，只有一页时写成 1。把它换成你真正要的页。再选一份 PDF，会换掉当前打开的那份。一次拖进好几个文件时，只打开第一个。这个工具不能一次拆整个文件夹。

A list such as 2-4, 11 copies page 2, page 3, page 4, then page 11, into that one PDF. The order is the order you type, not a sorted order, so 11, 2-4 puts page 11 first. A page can be repeated (1, 1) if you want it twice. English commas and Chinese commas both split the list. Extra commas are ignored. A hyphen, an en dash, or an em dash can mark a range. 5-2 is rejected instead of quietly reversed, so a typo does not flip a packet. A number outside the file, or a zero, is rejected too. The line under the box tells you how many pages will be copied once the list is valid.

像 2-4, 11 这样的写法，会把第 2、3、4 页，然后是第 11 页，放进那一个 PDF。顺序就是你输入的顺序，不是从小到大重排，所以 11, 2-4 会把第 11 页放在最前。同一页可以重复（1, 1），如果你需要它出现两次。英文逗号和中文逗号都可以用来分开各项。多出来的逗号会被忽略。连字符、短破折号或长破折号都可以表示范围。5-2 会被拒绝，而不是悄悄反过来，免得一个笔误把顺序倒过来。超出文件页数的数字，或 0，也会被拒绝。列表合法之后，输入框下面会告诉你将要复制多少页。

“One PDF per page” builds the zip in this browser. It is not assembled on a server. Inside the zip are only those PDFs, with no extra folder. Names are padded to the width of the page count so they sort: a 12-page file uses page-01.pdf, and a 100-page file uses page-001.pdf. A file with fewer than 10 pages is not zero-padded, so the first page is page-1.pdf. Copied pages keep their size, rotation, and real text, the same way Merge does. The bookmark outline, attachments, and document info are not rebuilt. A signature on the source does not become a valid signature of each new one-page file.

“每一页一个 PDF”的压缩包在这台浏览器里生成，不是在服务器上打包。包里只有这些 PDF，没有额外的文件夹。文件名按总页数的位数补齐，这样排序才对：12 页的文件用 page-01.pdf，100 页的文件用 page-001.pdf。不足 10 页时不补零，第一页就是 page-1.pdf。复制出来的页面会保留尺寸、旋转和真正的文字，和合并时一样。书签目录、附件和文档信息不会重建。源文件上的签名，不会变成每一份新的单页文件的有效签名。

One run covers at most 200 pages. Saving every page refuses a file longer than 200 pages outright. A longer file can still give you a range, as long as the pages you list are 200 or fewer: a 300-page handbook can spare pages 1-40, and you can run 41-80 as a second pass. Each file must be 80 MB or smaller. Password-protected and damaged PDFs are refused, and the password is not removed. Do not use Split to make a Word file, to OCR a scan, to burst a PDF into JPEG images, or to cut every 10 pages with one switch. There is no chunk-size control. Type the ranges yourself. To put several extracts back together, use Merge.

一次最多处理 200 页。选择每一页单独保存时，超过 200 页的文件会被直接拒绝。更长的文件仍然可以按范围抽取，只要你列出的页数不超过 200：一本 300 页的手册可以先抽 1-40，再跑一次 41-80。每个文件不超过 80 MB。有密码的和损坏的 PDF 会被拒绝，密码也不会被去掉。不要用拆分来做 Word、给扫描件做 OCR、把 PDF 拆成 JPEG，或指望一个开关就每 10 页切一块。这里没有“按固定页数分块”的选项。范围要自己写。如果要把抽出来的几段再拼回去，用合并。

StayFile copies the pages you name into a new PDF or a zip of new PDFs. It does not check rights in the source file. Encrypted PDFs are refused, and passwords are not removed. The bookmark outline, attachments, and document info are not copied. A digital signature does not remain valid on the pieces. Confirm the page index in a viewer before you send the result, because a printed footer can disagree with it. The original file on disk is not modified.

StayFile 把你点名的页面复制进一个新 PDF，或一组新 PDF 的压缩包。它不检查源文件的权利。加密的 PDF 会被拒绝，密码也不会被去掉。书签目录、附件和文档信息不会被复制。数字签名在拆开的文件上不再有效。发出结果之前，先在阅读器里核对页面序号，因为页脚印的数字可能和它不一致。磁盘上的原文件不会被修改。

The ceiling is 200 pages in one run, and 80 MB for the file you open. Saving every page needs the whole file to be within 200 pages. A range can take up to 200 pages from a longer file, in the order you type, including repeats. There is no control for “every 10 pages”, for a target file size, or for exporting page images. Build the zip or the PDF in this tab. Closing the tab drops the in-memory copy. It does not delete the download you saved.

一次最多 200 页，你打开的文件不超过 80 MB。每一页单独保存时，整份文件必须在 200 页以内。按范围抽取时，可以从更长的文件里取出最多 200 页，顺序按你的输入，也可以重复某一页。这里没有“每 10 页一块”、按目标大小拆分，或把页面导出成图片的选项。压缩包或 PDF 在这个标签页里生成。关掉标签页会丢掉内存中的副本。它不会删除你已经保存的下载。