Compare PDFs
在浏览器中使用「Compare PDFs」完成代码、数据与二维码相关任务；输入内容和结果默认在设备上处理。
使用步骤
如何使用
- Choose two PDF files within the size and page limits.
- Set the pixel tolerance and start the visual comparison.
- Review the page summary and difference previews; this compares rendered appearance, not text or document meaning.
Visual comparison only
PDF.js renders both documents page by page to bounded browser canvases. The result reports changed pixels and a heatmap-style preview; it does not compare text, metadata, structure, accessibility, or semantic meaning.
Thresholds and limits
The channel tolerance is adjustable from 0 to 64. Files are limited to 50 MB each, 100 pages each, 256 MB of working canvas memory, and 2,000 pixels on the longest rendered side. Pages with different sizes are compared over the shared canvas area and reported clearly.
Private local processing
Both files are read locally with PDF.js. No upload or server request is used for document data. Temporary object URLs, PDF documents, render canvases, and image buffers are released when a comparison is replaced, cleared, cancelled, or the page is left.
常见问题
你可能还想知道
Does this compare PDF text?
No. It compares the pixels produced by rendering each page. A text-aware or semantic comparison may reach a different conclusion.
What does tolerance mean?
It is the per-channel RGB difference ignored for a pixel. Higher values reduce sensitivity to anti-aliasing and tiny rendering changes.
Are my PDFs uploaded?
No. The files are opened and rendered in this browser with temporary local data.
Why are different page sizes reported?
The pages are rendered at a common bounded scale and compared over the shared area; the summary also reports the size mismatch.