Video2X
github.com
- Free plan
- Windows
- macOS
- Linux
- Self-hosted
7.6 Club score Visit github.com
What the maker publishes
- Acceleration
- Real-ESRGAN, Real-CUGAN, and RIFE run via ncnn and Vulkan. source
- Cloud use
- The project describes using Video2X on Google Colab, with free GPU sessions limited to 12 hours per session. source
- Colab
- The project offers a Google Colab notebook that can use free NVIDIA T4, L4, or A100 GPUs for up to 12 hours per session. source
- Containers
- Container images are available through GitHub Container Registry for deployment on Linux and macOS using Docker or Podman. source
- Container support
- Container images are available through GitHub Container Registry for deployment on Linux and macOS, using Docker or Podman. source
- Custom shaders
- It supports custom MPV compatible GLSL shaders. source
- Disk space
- The project says processing needs no additional disk space beyond space for the final output. source
- Documentation
- The documentation covers building, installing, using, and developing Video2X, and notes that some pages are under construction. source
- Founded
- 2018 source
- Hardware requirements
- Precompiled binaries require an AVX2 capable CPU and a Vulkan capable GPU. source
- Help and documentation
- The project links documentation for building, installing, using, and developing Video2X, plus a Telegram discussion group where users can ask developers questions. source
- Intended users
- The project describes its Colab option as useful for people who do not have a powerful GPU of their own. source
- Interpolation models
- It supports RIFE models through ncnn and Vulkan. source
- License
- Video2X is licensed under GNU AGPL version 3. source
- Linux
- Linux packages are listed for Arch Linux, and an AppImage is available for other distributions. source
- Linux availability
- Linux packages include Arch AUR options and a universal AppImage for other distributions. source
- Processing modes
- It supports video upscaling through filtering and frame interpolation. source
- Processing storage
- The project says processing requires no additional disk space beyond space for the final output. source
- Purpose
- Video2X is a machine learning based framework for video super resolution and frame interpolation. source
- Support
- The project points users to a Telegram discussion group to ask questions and talk with developers. source
- Supported models
- It supports Anime4K v4, Real-ESRGAN, Real-CUGAN, and RIFE models. source
- Upscaling models
- It supports Anime4K v4, Real-ESRGAN, Real-CUGAN, and custom MPV-compatible GLSL shaders. source
- Windows
- The project provides a Windows installer and a graphical interface. source
- Windows release
- The repository links a Windows installer and describes a GUI for Windows. source
Checked against the maker's pages on 7 Oct 2026.