Pixelify uses machine-learning models that run in the browser through supported inference runtimes and execution providers. The exact engine available can depend on browser, operating system, hardware, and model compatibility.
Pixelify model library
| Model / family | Purpose |
|---|---|
| Chopper | Background removal |
| Vitality | Image upscaling |
| Pixel-Dance | Image upscaling |
| Alpha-Satoru | Image upscaling |
Model weights may be downloaded to your browser and stored in its cache for reuse. This is separate from uploading your image for remote processing.
Execution engines
Depending on the device and browser, Pixelify may use WebGPU or a WebAssembly-based execution path. WebGPU can accelerate compatible workloads, while WASM can provide a broader compatibility fallback. Availability and performance are browser-dependent.
Improve performance
- Use a current version of Chrome, Edge, or another supported browser.
- Close memory-heavy tabs and applications before processing large images.
- Keep the browser tab active during model initialization and processing.
- Try a smaller source image or lower upscale factor if memory is constrained.
- On laptops, connect power and use a performance-oriented power mode when appropriate.
Some mobile devices have limited memory or incomplete WebGPU support. If the engine does not initialize, update your browser, retry with a smaller image, or use a desktop/laptop for demanding workflows.