Real-CUGAN vs Refocused
Real-CUGAN
github.comRefocused
refocused.aiSide by side
| Figure | Real-CUGAN | Refocused |
|---|---|---|
| Price from | Free | — |
| Free plan | Yes | — |
| Free trial | — | Yes |
| Open source | — | — |
| Web | Yes | — |
| Windows | Yes | Yes |
| macOS | Yes | Yes |
| Linux | Yes | Yes |
| Android | Yes | — |
| Self-hosted | Yes | — |
| Plans published | 1 | 2 |
| Spec score · AI Video Upscalers | 8.1 (#12) | 8.1 (#13) |
| Spec score · Video Upscaling Software | 8.0 (#17) | 8.0 (#19) |
| Support | The maker invites users to leave suggestions and requests in the repository issues. | Users can contact support by email or submit a support ticket, which requires sign-in. |
| Upscaling | It supports 2×, 3×, and 4× upscaling; the 1× model is described as still in training. | The Refocus model provides motion-aware 4× super-resolution, while the flagship workflow chains Refocus into Regenerate-Q. |
Plans
In their own words
Real-CUGAN
Real-CUGAN is a free AI super-resolution model designed for anime images. It enlarges images by 2×, 3×, or 4× and offers denoising strengths, a conservative mode, and adjustable alpha. The 1× model is still in training. The English README describes five model weights for 2× and three each for 3× and 4×, covering different enhancement strengths. The Windows configuration processes both images and video, and the repository includes a video inference script. A linked NCNN implementation provides Windows, Linux, and macOS executables for Intel, AMD, Nvidia, and Apple-Silicon GPUs using Vulkan; its portable package includes binaries and models and states that it needs no CUDA or PyTorch runtime. Browser options include a demo and a Web-CPU version, though the README says the CPU demo is slow and recommends smaller images. Windows cache and tile settings can trade speed for lower GPU-memory use, including modes intended for very large images. The project also lists PyTorch and VapourSynth use paths and describes its architecture as compatible with Waifu2x-CUNet. The maker invites suggestions and requests through repository issues.
Refocused
Refocused is desktop software for enhancing and restoring video with AI models that process footage on the user's machine. Its seven model families cover super-resolution, diffusion restoration, interpolation, denoising, depth estimation, and segmentation. The Refocus model provides motion-aware 4× super-resolution, and a flagship workflow chains it with Regenerate-Q. In the Refine workspace, users can apply tracked object masks, select by depth threshold, reprocess specific regions, and create synthetic depth-of-field focus. Compare offers synchronized A/B playback and up to 8× zoom for inspecting frames. Output can reach 8K and includes 10-bit ProRes, HDR, and image sequences; listed formats include H.265/H.264, ProRes, VP9, AV1, PNG, JPEG, and TIFF. Processing works offline after installation and activation, though license validation and model updates periodically need internet access. A free trial includes full processing and all models, with watermarked output. The Full license is $64.99 as a one-time purchase and includes unlimited local processing, up to five devices, and one year of updates.