SeedVR2 Video Upscaler (ComfyUI) at a glance
Checked 16 Oct 2026Runs ByteDance's SeedVR2 3B or 7B restoration model as ComfyUI nodes or a standalone multi-GPU CLI that batch-processes folders of images or videos. It is generative, so it can add detail like Gigapixel's generative models. Code and official weights are Apache-2.0. It is heavy: 8GB VRAM only with GGUF quantisation and offloading, 24GB+ for full precision. It needs a Python/PyTorch setup; there is no simple installer. No face-specific model. Last commit Dec 2025.
RepoQuarry. “SeedVR2 Video Upscaler (ComfyUI).” Checked 16 Oct 2026. https://repoquarry.dev/projects/seedvr2-comfyuiHow much of Topaz Gigapixel does it replace?
3 of 6 jobs. Closest open match to Gigapixel's generative upscaling, at a cost of high VRAM and a technical setup.
For face recovery on blurry or low-res faces, upscaling art, Photoshop and Lightroom plugin, see other Topaz Gigapixel alternatives. You may also need to build an interface, an API server and hosting yourself.
Facts, with sources
Every answer links to where we found it. We read the documentation; we did not install it.
Official release of SeedVR2 for ComfyUI that enables high-quality video and image upscaling.github.com · ComfyUI-SeedVR2_VideoUpscaler#readme ↗ today
The code in this repository is released under the Apache 2.0 license as found in the LICENSE file.github.com · ComfyUI-SeedVR2_VideoUpscaler#-license ↗ today
Minimal VRAM (8GB or less): Use GGUF Q4_K_M models with BlockSwap and VAE tiling enabledgithub.com · ComfyUI-SeedVR2_VideoUpscaler#hardware ↗ today
Apple Silicon Support: Full MPS (Metal Performance Shaders) support for Apple Silicon Macsgithub.com · ComfyUI-SeedVR2_VideoUpscaler#-run-as-standalone-cli ↗ today
Models will be automatically downloaded on first use and saved to ComfyUI/models/SEEDVR2.github.com · ComfyUI-SeedVR2_VideoUpscaler#-installation ↗ today