Oversized PDFs block email attachments, slow down uploads, and eat through cloud storage quotas. The free Compress PDF tool on SapienKit shrinks your files by re-encoding embedded images at a lower quality setting you control, all without uploading a single byte to any server. Everything runs in your browser — no account, no watermark, no daily cap. Drop your PDF in, adjust the quality slider, and download a leaner version in seconds.
The compression pipeline works by extracting every embedded image stream from the PDF, decoding it into raw pixel data via the Canvas API, then re-encoding it as a JPEG at the quality level you choose on the slider. The re-encoded image is written back into the PDF structure using pdf-lib, replacing the original stream. Text, vector graphics, fonts, and annotations are left completely untouched — only raster images are affected. This approach is especially effective because images are almost always the largest payload inside a PDF. A 50-page report with high-resolution photos can easily drop from 40 MB to 8 MB at a quality setting of 60, while the text on every page remains crisp and fully searchable.
The practical use cases span nearly every profession. A job applicant needs to email a portfolio PDF under a 10 MB attachment limit. A sales representative wants to attach a product catalog to a CRM record capped at 5 MB. An accountant archives hundreds of scanned invoices and needs to cut storage costs in half. A student submits a research paper with embedded figures to an LMS that rejects files over 20 MB. A government clerk compresses planning documents before publishing them on a public records portal with bandwidth constraints.
More niche workflows benefit just as much. Architects can shrink PDF plan sets full of high-DPI renderings before emailing them to contractors. Photographers compressing proof sheets can reduce gallery PDFs from 100 MB to under 15 MB. Nonprofit administrators preparing grant applications can ensure every appendix attachment fits within the submission portal's strict file-size rules. Insurance adjusters can compress claim packets stuffed with photographic evidence so they transfer faster over spotty mobile connections.
Cloud-based PDF compression tools require you to upload your document, wait for server-side processing, and then download the result. Free tiers typically limit the number of compressions per day, restrict file size, or add a watermark. The more serious concern is privacy: your file travels to and is processed on a third-party server, which may conflict with confidentiality agreements, HIPAA obligations, or GDPR requirements. SapienKit's browser-based approach eliminates all of these issues. Compression starts instantly on your own hardware, scales with your CPU rather than server queue depth, and keeps your data firmly on your device.
The tool handles PDF versions up to 2.0 and correctly processes files with compressed object streams and cross-reference streams. Vector graphics — diagrams, charts, logos drawn with paths — pass through untouched and remain perfectly sharp at any zoom level. Embedded fonts, including subset-embedded and CJK fonts, are preserved byte-for-byte. Hyperlinks, form fields, and annotations survive compression intact. The maximum file size you can process depends on your browser's available RAM; most modern machines handle PDFs up to 150-200 MB comfortably, though extremely image-dense files above 300 MB may cause slowdowns on devices with limited memory.
A few tips for the best results. Scanned PDFs benefit the most from compression because every page is essentially a full-page image — expect 50-80% size reduction at moderate quality. Digitally created, text-heavy PDFs with few images will shrink only marginally, since text and vector data are already compact. Use the quality slider to find the sweet spot: a setting around 60-70 gives a strong size reduction with minimal visible difference; going below 40 will produce noticeable artifacts in photographic images. If your PDF is password-protected, decrypt it first with the Unlock PDF tool — the compressor cannot read encrypted image streams. And remember, compression is a one-way trade-off for images: re-encoding a JPEG that was already compressed will degrade it further, so avoid compressing the same file multiple times.
Privacy is central to how this tool works. Your PDF is read into browser memory through the File API and processed entirely on your device. No file data, metadata, or telemetry is sent over the network. The tool functions fully offline once the page has loaded — you can disconnect your internet and still compress a PDF. When you close the tab, all data is purged from memory. This design is inherently GDPR-friendly, HIPAA-compatible, and suitable for any workflow where documents must never leave a controlled environment.
The tool extracts every embedded image from the PDF, decodes it into raw pixels using the Canvas API, re-encodes it as a JPEG at your chosen quality level, and writes it back into the PDF structure via pdf-lib. Text, vectors, and fonts are untouched. Everything runs in your browser — no server involved.
It depends on image content. A PDF full of high-resolution photos or scans can shrink 50-80% at a quality setting of 60. Text-only PDFs with few images may only shrink 5-10%, since text and vector data are already compact.
No. Compression only targets raster images (photos, scans). Text remains fully searchable and sharp. Vector graphics like charts, logos, and diagrams pass through untouched and stay crisp at any zoom level.
Yes, and scanned PDFs benefit the most because every page is essentially a full-page image. A 50 MB scanned document can easily drop to 10-15 MB at moderate quality. This is the single best use case for the tool.
It controls the JPEG re-encoding quality for embedded images, from 1 (maximum compression, lowest quality) to 100 (minimal compression, highest quality). A setting of 60-70 gives strong size reduction with minimal visible difference. Below 40, photographic images may show noticeable artifacts.
There is no hard limit. Processing runs in browser memory, so it depends on your device's RAM. Most modern computers handle PDFs up to 150-200 MB comfortably. Extremely image-dense files above 300 MB may cause slowdowns on machines with limited memory.
The tool cannot read encrypted image streams. Remove the password first using the [Unlock PDF](/pdf/protect) tool, then compress the decrypted file. This is a limitation shared by all client-side PDF processing libraries.
For quick, private compression of image-heavy files, yes. Desktop PDF editors offer more granular controls like font subsetting and object stream optimization, but they often require a paid subscription. SapienKit is free, instant, and keeps your files on your device.
You can, but it is not recommended. Each round re-encodes the already-compressed images, which degrades quality further without much additional size reduction. Compress once at the right quality setting for the best result.
Nowhere. Your PDF is read into browser memory and processed entirely on your device. No file data, metadata, or telemetry is sent to any server. The tool works offline once loaded, and closing the tab purges everything from memory.