Windows desktop AI application for upscaling, enhancing, and denoising photos and videos completely offline.
Enhance photo details, upscale videos, and reduce noise offline using your local DirectX 12 GPU.
QualityScaler is a local Windows desktop application designed to upscale, de-noise, and enhance images and videos using deep learning models. It runs entirely on your hardware without transmitting data to external servers, making it ideal for privacy-conscious photographers and video editors.
The app processes raster images and various video formats through AI architectures like BSRGAN and Real-ESRGAN to boost resolution and clear up compression artifacts. It includes automatic image tiling to prevent GPU memory exhaustion and lets you blend upscaled output with original files using adjustable interpolation levels.
It suits photographers, archivists, and videographers who need local batch upscaling without cloud subscription fees or internet access. It is not suitable for users on macOS or Linux, or those without a dedicated DirectX 12 graphics card.
Splits large photos into smaller tiles during processing to avoid running out of GPU memory.
Runs entirely on your local graphics hardware without requiring an internet connection or uploading files.
Blends original images and upscaled outputs at low, medium, or high levels to maintain natural detail.
Transfers metadata from original photo files directly to upscaled outputs using exiftool.
QualityScaler is free open-source software under the MIT license. There are no paid cloud tiers, so your only cost is the hardware and electricity needed to run processing locally.
767 open remote AI-training roles, updated November 14, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
These reviewed AI tools do the same job instantly — from our scored directory.
No, all image and video processing happens locally on your computer's graphics processor.
Yes, it uses ONNX Runtime DirectML to support DirectX 12 compatible GPUs from AMD, Intel, and Nvidia.
See Zekai first in Google