Our take
Activerecent releases- Activity
- 45
- Community
- 20
- Issues
- 47
- Pull requests
- 20
+83 stars in the last 30 days
This is an actively released, capable multi-platform video-enhancement GUI, but its maintenance is heavily concentrated in one contributor and successful use depends on compatible GPU software and hardware.
Good fit if
- You want a desktop workflow for interpolation, upscaling, decompression, or denoising on Windows, Linux, or macOS.
- Your GPU supports one of the documented backends, and you are comfortable selecting models and troubleshooting inference setup.
- You can comply with the AGPL-3.0 license and want a project that has continued to publish releases.
Look elsewhere if
- You need a project with broadly shared maintenance, since most contributions come from one person.
- Your machine lacks a compatible Vulkan, CUDA, ROCm, or NVIDIA RTX setup for the backend you intend to use.
- You need a documented automated test suite or a workflow that does not depend on model inference and GPU configuration.
All health signals
| Last commit | 2026-09-02 (2 months ago) |
|---|---|
| Commits, last 90 days | 1 |
| Releases, last 12 months | 4 (latest RVE-2.4.1, 2026-01-02) |
| Contributors | 7 (top contributor: 99% of commits) |
| Issues closed, last 90 days | 1 |
| Pull requests merged, last 90 days | 0 |
| Project age | 3 years |
Checked on 2026-10-20 with the GitHub API.
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
2,404Forks
168Change since Nov 11, 2026: 0 stars·0 forks
Overview
REAL Video Enhancer is a cross-platform desktop app for video interpolation, upscaling, decompression, and denoising. It builds on models such as RIFE from Practical-RIFE, alongside GMFSS and IFRNet for interpolation, and provides TensorRT, PyTorch, and NCNN inference backends.
The app is intended to make these model-based video enhancements accessible through a GUI on Windows, Linux, and macOS, rather than requiring users to assemble a separate processing workflow. Its usefulness depends on having compatible hardware and installing or selecting a suitable backend.
Key Features
- Interpolate frames with RIFE, GMFSS, and IFRNet models.
- Upscale video with supported models including Real-ESRGAN-related and SPAN models.
- Decompress and denoise using models such as DeH264, DRUnet, and DnCNN.
- Choose TensorRT for NVIDIA RTX GPUs, PyTorch for supported CUDA or ROCm hardware, or NCNN for Vulkan 1.3 devices.
- Detect scene changes to help preserve sharp transitions.
- Preview the latest rendered frame during processing.
- Build distributable apps using PyInstaller, cx_Freeze, or experimental Nuitka support.
Use Cases
- Video editors can increase frame rates or resolution for footage that needs smoother playback or larger output dimensions.
- Anime and animation enthusiasts can try the listed interpolation and upscale models on their own clips.
- Linux, Windows, or macOS users seeking a GUI can apply multiple enhancement stages without building a custom inference pipeline.
- GPU owners can compare available inference backends and choose one compatible with their hardware.
What you need
Detected in the repository
- Python (from requirements.txt)
- Automated checks on GitHub Actions
License in plain words
AGPL-3.0network copyleft
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the license notice
- Share your changes: yes, the full source, also when people only use it over a network (e.g. as a hosted service)
- Includes an explicit patent grant from the contributors:
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
Clone the repository with its submodules:
git clone --recurse-submodules https://github.com/TNTwise/REAL-Video-Enhancer
See the README for stable-branch instructions, downloads, build steps, model details, and backend guidance.
Alternatives
- paper2gui: paper2gui bundles video enhancement alongside other AI tools, while REAL-Video-Enhancer focuses on video processing such as interpolation, upscaling, and denoising.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| REAL-Video-Enhancer | Python | AGPL-3.0 | 2.3k | Active |
| paper2gui | Jupyter Notebook | MIT | 10.7k | Inactive |
Considerations
- Video inference is hardware-sensitive. The README recommends a Vulkan 1.3-capable device at minimum, and describes TensorRT as targeting NVIDIA RTX GPUs. VRAM limits can cause TensorRT optimization or inference to fail.
- Backend setup relies on portable Python and PIP, which the project notes can sometimes fail to install or detect correctly. ROCm and Vulkan errors may require separate troubleshooting.
- The project has recent releases, but recent commit and issue activity is limited, and contribution activity is highly concentrated in one person. The README lists no tests, so adopters should validate output and packaging against their own workloads.
- The project is licensed under AGPL-3.0. Review its obligations before distributing modified versions or incorporating it into a larger product.
- The README warns that the Windows executable may be flagged as a trojan by some scanners, attributing this to PyInstaller false positives. Verify downloads through trusted project channels.
Found this useful?
Share it with someone who would like REAL-Video-Enhancer.