Convert 1,000 Photos in a Minute — Without Uploading a Single One

A thousand photos. Under a minute. Zero files uploaded anywhere. That’s the headline, and the rest of this post is why it’s possible — and why the “online converter” you’ve been using keeps getting in your way.

You’ve probably hit one of these walls:

• “Maximum 20 files per batch.” Upload 20. Wait. Download. Repeat 50 times.
• “100 MB limit on free tier.” Three DSLR photos and you’re done.
• “Upgrade to batch convert.” $12/month for a feature that shouldn’t cost anything.
• “Files deleted after 24 hours.” Meaning they were on a stranger’s server for those 24 hours.

These aren’t technical limitations. They’re business-model artifacts. The converter runs on someone else’s servers, every image you convert costs them CPU time and bandwidth, and the caps exist to push you toward a paid plan.

There’s a better way. Run the converter in your own browser, on your own machine, with no upload at all. It’s faster, it’s free, and your photos never go anywhere.

The honest math on 1,000 photos

A typical smartphone photo is about 3 MB. A 1,000-photo vacation folder is around 3 GB.

Uploading to an online converter:

1. Upload 3 GB over your home connection. On a typical 50 Mbps upload, that’s roughly 9 minutes before any conversion even begins.
2. Wait in a queue — free tiers usually deprioritize large jobs.
3. The server converts the files.
4. Download the results (~2 GB). Another 5 minutes.
5. Total: 14+ minutes, 5 GB of your ISP bandwidth, and your photos touched an external server.

Converting in your browser (WebP target):

1. Drop 1,000 files on the page. They’re read from your SSD at disk speed — takes a couple of seconds.
2. Workers start encoding immediately. On a 10-core laptop, that’s ~30 photos/sec throughput.
3. 1,000 photos finish in roughly 33 seconds.
4. A ZIP is assembled on your device and saved.
5. Total: under a minute, zero upload, nothing left your machine.

The browser is faster. That’s not a typo.

Pick AVIF and those numbers stretch — a lot. The AV1 encoder underneath AVIF is roughly 5–20× slower than WebP per photo. A 12 MP AVIF encode lands around 3–8 seconds even with every core in the fight, so 1,000 photos to AVIF is more like a coffee break — 10–30 minutes, not 33 seconds. Still faster than the upload dance, still private, but honest: WebP when you’re trying to get out the door, AVIF when you want the smallest files and can leave the tab running. WebP vs AVIF in 2026 has the full trade-off.

Why your browser can actually handle this

Every core on your machine, already paid for, already sitting idle while you wait for an upload bar. The work here is a native codec crunching pixels — the same kind of work a desktop app does — and modern browsers expose enough of the machine to run that work at close-to-native speed.

Each tab gets up to 4 GB of working memory. Web Workers give you true parallelism across every logical core. A WebAssembly sandbox lets a native codec stack run without leaving the page. So we compiled a real image pipeline — decoders, encoders, metadata, the whole kit — into a binary your browser caches on first visit and then keeps forever. Ten cores means ten encoders running in parallel. That’s where the 30 photos/sec comes from.

How to run it

1. Open scizone.dev/bulk-webp-converter in any desktop browser — Chrome, Firefox, Safari, Edge, or Opera.
2. Drag your photo folder onto the drop zone. Mixed formats (JPG, PNG, HEIC, TIFF) and subfolders all work in a single pass. Non-image files like .DS_Store pass through unchanged.
3. Watch the live counter. The UI stays fully responsive — scrolling, clicking, opening a new tab all work fine. Encoding runs in background workers, not on the main thread.
4. Download the ZIP. The output preserves your original folder structure, so extracting gives you a clean WebP mirror of your library.

A note on RAM and very large batches

Only the files currently being processed are in memory at any time — the rest are queued. The worker pool also recycles each worker after 32 jobs to keep memory from accumulating across long runs.

For very large batches (5,000+ images), we’d recommend splitting into chunks of about 1,500. The browser can technically handle more, but it shares memory with other tabs and background processes, so chunks reduce the risk of bumping into limits mid-batch.

Does it really work offline?

Yes — and you can prove it. Once the page has loaded, open DevTools → Network and drop a batch of files. You’ll see zero network activity during the entire conversion. The page is also a PWA, so cutting your Wi-Fi mid-run won’t interrupt anything.

If you want offline-by-default, install SciZone as a PWA from your browser’s address bar. It launches like a native app after that, completely offline, every time.

What happens to EXIF and GPS data?

Everything is preserved automatically. Capture timestamps, GPS coordinates, camera model, ICC color profiles — all of it copies onto the WebP output.

This matters a lot for large batches. Your photo library software (Apple Photos, Google Photos, Lightroom, digiKam) uses this metadata to organize photos by time and location. Lose it, and a carefully organized archive becomes a flat pile of files with no dates.

If you want to strip EXIF — say, for a batch you’re posting publicly — add a pass through ExifTool after conversion. The default is to keep everything.

Verify it yourself

Privacy claims are easy to make and hard to trust. Here’s how to independently confirm this one:

1. Open the bulk converter in Chrome.
2. Open DevTools → Network. Enable “Preserve log.”
3. Drop 10 photos onto the page.
4. Filter the Network tab by POST, PUT, or “upload.”

You’ll see nothing related to your image data. The only requests after page load are for cached static assets from scizone.dev.

The bottom line

In 2026, there’s no reason to accept batch caps, upload waits, or privacy compromises for photo conversion. Your browser can do this work directly — faster than the upload path, completely free, with no ceiling on batch size. Drop your photos on the bulk converter and watch how fast it actually goes.