Recadrer PDF en ligne gratuit – rogner pages, marges

Open your .pdf by clicking the Open PDF (or drag & drop) area, then use the blue crop handles drawn on the page preview to drag corners or edges. For numeric precision switch to the Units dropdown and enter margin values in mm/in/pt, or use the Page range box (example: 1-3,5) before you press Crop. Check the Apply to all pages checkbox if you want the same crop on every page, then click Download PDF to save the resulting .pdf to your browser Downloads folder. All processing happens locally in your browser, so there is no file upload step.

This tool accepts PDF files only (.pdf) as input and outputs a cropped .pdf file. Desktop Chrome and Firefox handle files up to about 50 MB and 200 pages comfortably; on mobile Safari or low-memory devices you should keep files under ~20 MB to avoid browser crashes. Extremely complex PDFs (many high-res images or layered transparency) may exceed those limits and trigger out-of-memory errors, in which case split the PDF into smaller chunks or use a desktop app like Acrobat or Ghostscript.

The application processes files locally in your browser using the File API and WebAssembly, so your .pdf is not uploaded to any remote server during the crop operation. You can verify this by opening DevTools → Network and watching that no outgoing POST/PUT occurs when you click Crop or Download. Note that browser behavior (permissions and temporary memory) still applies — nothing leaves your device unless you explicitly use a separate share/export option in your browser.

Yes: click the Auto Crop button in the toolbar to run a margin detection pass that analyzes whitespace by a tolerance slider expressed in pixels (default 10 px). The preview shows the detected crop on each page and you can tick Apply to all pages or limit it with the Page range input (for example 1-5,8). Auto Crop modifies the PDF CropBox/TrimBox without re-encoding page images, so visual fidelity is preserved unless you enable rasterization under Advanced options. If detection misses some pages, tweak the tolerance or manually adjust handles on those specific pages.

Yes — use the Page range field before applying the crop; it accepts single pages and ranges separated by commas (for example: 1-3,5,7). Alternatively, select a page in the left-hand thumbnail strip, set the crop for that thumbnail and click Apply to selection to change only the chosen pages. The Apply to all pages checkbox will override your selection if checked, so be sure it's off when editing individual pages. After cropping the selected pages, use Download PDF to save a single .pdf containing the mixed cropped and uncropped pages.

By default the tool adjusts the PDF page boxes (CropBox/TrimBox) which does not re-encode images, so visual quality stays identical and file size remains very similar. If you enable the Flatten and Optimize option the pages are rasterized and re-compressed; the optimizer offers a JPEG quality slider (10–100) and choosing lower quality will reduce file size but introduce lossy artifacts. Rasterization can easily increase file size by 2–10× for heavily vector PDFs with embedded high-res scans, so only use it when you need flattened output for legacy readers.

If the PDF is encrypted you will see an Encrypted PDF dialog asking for the owner/user password; cropping cannot proceed without entering the correct password. For scanned PDFs (images on pages) cropping behaves the same — it modifies the page boundary but does not perform OCR or alter embedded image DPI. If you need editable text from a scanned PDF, run OCR in Acrobat or a dedicated OCR tool before or after cropping; cropping scanned pages does not improve text recognition by itself.

Mobile browsers — particularly iOS Safari — have tighter memory limits and single-process WebKit constraints, so large files (>20 MB or many high-resolution images) can cause the tab to become unresponsive or crash. Workarounds: use desktop Chrome/Firefox for files over ~50 MB, split the PDF into smaller files, close other apps/tabs to free RAM, or reduce image-heavy pages first in a desktop tool. Also note that drag & drop is limited on iOS; use the Choose File button instead to open PDFs from Files.

Chrome and recent Firefox support multi-threaded WebAssembly which makes cropping faster and more stable on large documents; they also allow drag & drop and long-running scripts. Safari (desktop and iOS) restricts background threads and has lower memory ceilings, so operations may be slower or hit WebKit limits — use Chrome for heavy batches. Another practical quirk: in Firefox the browser may append a suffix to duplicate download names; if that happens right-click the Download button and use Save Link As to control the filename.

No signup or payment is required to use the core cropping features — the tool is free to use and does not add watermarks to the output .pdf. After you click Download PDF the cropped file is saved to your browser's Downloads folder (or the mobile Files app on iOS when prompted). Keep in mind that optional advanced features like rasterization/optimization are integrated into the same free interface; there is currently no paid tier or hidden trial gating the basic crop functionality.

Choose Units from the crop toolbar and switch to mm or pt, then enter exact values into the Top/Bottom/Left/Right numeric fields; 1 in = 25.4 mm = 72 pt, so a 3 mm bleed equals 8.5039 pt (you can round to 8.5 pt). If you want to resize the page to a standard sheet use the Resize page to dropdown and pick A4 or Letter, or enter custom page width/height in mm/pt before applying. After setting the numeric values preview each page to confirm the crop matches printer-required bleed before exporting.

Use Crop PDF Online Free for quick, single-file edits: removing white margins, trimming pages, or basic page-box resizing without installing software. Acrobat Pro is better when you need advanced box control (CropBox/ArtBox/TrimBox), batch actions across hundreds of files, form or redaction features, or built-in OCR. Photoshop rasterizes pages and is appropriate when you need pixel-level image edits but expect larger files; Ghostscript or qpdf are preferable for automated command-line batch workflows and lossless box changes on large corpora.