Video2X
A local AI video upscaler increases resolution and interpolates frames on Windows and Linux using a Vulkan-capable GPU.
Alternative to:
Video2X processes video on your own hardware to increase resolution or add frames for smoother motion. It's for people improving existing footage, with several processing methods suited to anime and other video.
- Video upscaling. Choose Anime4K, Real-ESRGAN or Real-CUGAN to increase video resolution.
- Frame interpolation. Use RIFE to generate intermediate frames and increase a video's frame rate.
- Custom filters. Apply MPV-compatible GLSL shaders alongside the supported upscaling methods.
Video2X is open source under AGPL-3.0 and runs on Windows and Linux. Windows has a graphical interface; Docker and Podman images are available for Linux and macOS. Neural models use ncnn with Vulkan acceleration. A compatible NVIDIA, AMD or Intel GPU must support Vulkan, and precompiled binaries require an AVX2-capable CPU.
Local processing runs on your machine. An optional Google Colab notebook runs the workload on Google's servers instead. The processing pipeline doesn't create intermediate files that require extra disk space; it needs storage for the final output.
Logotyper
