← Back to all tools

TOOL · 100 · AI

AI Image Upscaler (2×)

Enlarge photos 2× with a real Swin2SR super-resolution neural network — sharper than plain resize. In-browser, no signup. First run ~50 MB.

In short
AI Image Upscaler enlarges photos 2× with a real Swin2SR super-resolution neural network — running in your browser via transformers.js. Produces genuinely reconstructed detail rather than a plain stretch.

Runs in your browser after a one-time model download. Your files stay on your device.

Drop files here or click to browse

JPG, PNG or WebP · under 512px works best

No signup · No watermark · Files never leave your device · Up to ~300 MB recommended

How it works

  1. Step · 01

    Drop a JPG, PNG or WebP image (best under 512px on the long side for speed).

  2. Step · 02

    First run downloads the open-source Swin2SR-x2 super-resolution model (~50 MB) — cached after that.

  3. Step · 03

    The neural network runs locally in your browser and downloads a sharp 2× PNG.

Related tools

Frequently asked questions

Which model does it use?
Xenova/swin2SR-classical-sr-x2-64, an open-source Swin Transformer for image restoration published at ECCV 2022.
Why is input capped at 512 px?
Swin2SR is compute-heavy — inputs are auto-resized down to 512 px on the long side so the browser can finish in a reasonable time. The output is then 2× that.
How does this compare to the plain Image Upscaler?
The other upscaler uses fast multi-pass canvas scaling — no download, no AI, works instantly. This one uses a real neural network that reconstructs detail, at the cost of a 50 MB one-time model download and slower processing.
Is my image uploaded?
No. The upscaling runs entirely on your device after the one-time model download.
Are my files uploaded to a server?
No. Files are processed directly in your browser using JavaScript and WebAssembly. Nothing is uploaded to our servers, so your documents never leave your device.
Does it work on mobile?
Yes. The tools run in any modern browser on desktop, tablet, or mobile (Chrome, Edge, Safari, Firefox). Very large files may be slower on low-end phones.