Compress JPG, PNG, WEBP — In Your Browser
Drop images below. Compression runs locally using the Canvas API — nothing is uploaded, no signup required, no file size or batch limits.
JPG, PNG, WEBP · Multiple files OK · Any size
orHow to compress images
How image compression works
There are two types of image compression. Lossy compression (used by JPG and WEBP) discards image data that the human visual system is least sensitive to — subtle color variation in uniform areas, fine texture in shadows, and high-frequency detail that blurs at normal viewing distances.
At quality 80, most photos look indistinguishable from the original but are 50–70% smaller. The quality slider controls how aggressively data is discarded. Below quality 50, visible artifacts appear — blocky shadows, color banding, blurry edges around text.
Lossless compression (used by PNG) finds patterns in the pixel data and encodes them more efficiently, without discarding any information. The output is bit-for-bit identical to the original. Lossless produces smaller files than uncompressed, but larger than lossy at the same visual quality.
Photos and camera images: use JPG. Lossy compression works exceptionally well on photographs because the discarded detail is in the exact frequency ranges the eye doesn't notice. A quality-80 JPG of a landscape is typically 5× smaller than a lossless PNG with no perceptible difference.
Screenshots, UI captures, logos, diagrams: use PNG. These contain sharp edges, text, and large solid-color regions — exactly the kind of detail lossy compression smears. PNG preserves every pixel exactly.
When file size matters above all else: convert to JPG with quality 75–85. This typically beats any lossless format by a significant margin. If the image has transparency, the "Convert to JPG" option fills transparent areas with white before encoding.
When to compress images
- Web publishing: Large images are one of the biggest causes of slow page loads. Compressed images improve page speed scores and reduce bandwidth for both you and your visitors.
- Email attachments: Most email services cap attachment sizes at 10–25MB. Compressing images lets you send more in a single message without triggering size limits.
- Cloud storage: Batch-compressing a photo library can cut storage by 50–70%, preventing plan overages or freeing space for more files.
- Form and portal uploads: Many document portals, HR systems, and e-commerce platforms impose file size limits. Compression gets images under those thresholds without re-shooting.
- Social media: Most platforms re-compress your images anyway. Pre-compressing gives you control over what gets discarded before the platform's algorithm takes over.
- Archiving a large photo library: Batch compression at quality 85 can halve storage requirements while keeping images visually identical for everyday viewing.
Quality settings reference
| Quality range | Best for | Visual quality | File size reduction |
|---|---|---|---|
| 90–100 | Archival copies, print preparation | Indistinguishable from original | 20–40% smaller |
| 75–90 | Web use — most common choice | No visible difference at normal viewing | 50–70% smaller |
| 50–75 | Thumbnails, aggressive compression | Slight softness visible on close inspection | 70–85% smaller |
| Below 50 | Maximum compression only | Visible artifacts, blocky shadows | 85–95% smaller |
Highlighted row (75–90) is the recommended range for most web and sharing use cases.
How this compares to TinyPNG and similar tools
TinyPNG dominates image compression. Here's an honest side-by-side.
| Feature | This tool | TinyPNG (free tier) |
|---|---|---|
| Where compression runs | Your browser (Canvas API) | TinyPNG's servers |
| Files uploaded to server? | Never | Yes |
| Max file size | No limit (browser memory) | 5MB per file |
| Max batch size | No limit | 20 images |
| Works offline? | Yes (after page loads) | No |
| Real-time preview? | Yes | No |
| Compression algorithm | Browser Canvas encoder | mozjpeg + zopfli (more aggressive) |
| Typical file size reduction | 50–70% at quality 80 | 60–80% (better algorithm) |
Is this safe for sensitive images?
Yes — and you can verify it yourself. This tool never uploads your images to any server. All compression happens inside your browser using the Canvas API, a standard browser feature available in Chrome, Firefox, Safari, and Edge.
When you drop an image, your browser reads it using the File API — this is local disk access, not an upload. The image data is decoded and drawn to an in-memory canvas element. The canvas re-encodes the data as a compressed blob at your chosen quality setting. That blob is made available for download via a temporary object URL, which is revoked when you're done. Nothing leaves your device.
To verify: open DevTools (F12) → Network tab → drop in some images and click "Compress." You'll see no outbound requests carrying your image data. The only network activity is loading the page itself on first visit.
Offline test: load this page, then disconnect your WiFi or enable airplane mode. The tool continues working perfectly — conclusive proof that it doesn't depend on a server.
This matters when compressing images that contain personal information: scanned IDs, medical photos, financial documents, confidential work images, or anything you'd rather not hand to a third party. The Canvas API approach means there's no server to be breached, no retention policy, and no terms of service that might allow use of your image data. See our full privacy policy for everything we do (and don't) collect.