Comprimir PDF
Reduza o tamanho dos seus arquivos mantendo a qualidade visual. 100% local no seu navegador.
墨观 PDF 的合并、拆分、压缩、加密等基础工具在你的浏览器里运行，处理过程中文件不上传服务器，断网也能用。
Click to select or drop PDFs here
Batch compression supported · processed entirely in your browser, never uploaded
"Compress" here covers two different operations. Standard compression only tidies up the file's structure — text stays selectable and searchable, and it typically saves 10–30%. Strong compression rasterizes every page into an image: the file usually gets much smaller, but the text can't be selected, Ctrl+F finds nothing, and editing is over. If a contract has to stay searchable in your archive, don't touch strong. If a scanned deck keeps bouncing off a 10MB upload limit, rasterizing is exactly what you want.
Both run in your browser (pdf-lib and pdf.js), so the file never leaves your machine. You don't have to take my word for it: press F12, open the Network panel, tick "Preserve log", and watch — from picking the file to downloading the result, no upload request should appear. A blunter test: disconnect from the internet first, then compress. It still works. The AI features here are a separate path that does go out, sending only locally extracted plain text to DeepSeek. Compression never touches it.
FAQ
Why did my PDF barely shrink — or get bigger — after compression?
Text-heavy PDFs have little fat to trim. Standard compression only reorganizes the file structure, so on a plain text layout you'll often land well under the 10–30% the tool advertises — a couple of percent is normal. Strong compression can backfire outright: rasterizing crisp vector text into page images frequently yields a file larger than the original. Compression pays off when scans or high-resolution images make up the bulk of the file.
Can I still search and copy text in a PDF after compressing it?
With standard compression, yes — it optimizes at the structure level only, the text layer survives, and Ctrl+F and copy-paste keep working. With strong compression, no. Each page is redrawn as an image, so no text objects remain: nothing to search, nothing to select, nothing to edit. For a contract or paper headed into an archive you'll need to search later, stay on standard.
Does the 50MB PDF compression limit go up with Pro?
It doesn't — the code sets 50MB for free and Pro alike on this tool. Strong compression rasterizes whole pages, and peak memory runs several times the file size, so the ceiling is really about what a browser tab can survive. Some other tools on this site do get a higher cap with Pro; compression is the exception, and I'd rather say that here than have you find out after subscribing.
Does compressing a PDF break form fields, links, or digital signatures?
Strong compression definitely does: pages become images, so fillable fields, clickable links, and the text layer are gone — only the visual appearance survives. Standard compression keeps the text layer, but it still rewrites the file structure, which can invalidate a digital signature that depends on the original bytes. If a document is signed or meant to be filled in, open the compressed copy and check it before sending it anywhere.
Is browser-based PDF compression weaker than server-side compressors?
Often, yes. Server-side services can throw heavier image resampling and font subsetting at a file and have no browser memory ceiling to respect. Running in a tab, this tool has fewer moves available, and on some files a hosted service will beat it. The trade is simple: they get a copy of your document, I don't. If the PDF isn't sensitive and standard compression did nothing, trying a server-side tool is reasonable.