Skip to content
SPECIFICTION
Get listed

DiffBIR

github.com

Free plan WebmacOSLinuxSelf-hosted
Visit github.com
Spec score 8.1 #43 of 99 in Image Upscaling Software

Spec sheet

Purpose
DiffBIR is a blind image restoration pipeline that uses a generative diffusion prior.Read 12 Oct 2026 · github.com
Deployment
The instructions clone the repository, create a Python 3.10 Conda environment, install requirements, and run inference locally.Read 12 Oct 2026 · github.com
Support
The README directs questions to [email protected].Read 12 Oct 2026 · github.com
License
The project is released under the Apache 2.0 license.Read 12 Oct 2026 · github.com
Headquarters
ChinaRead 28 Sep 2026 · github.com
Founded
2023Read 28 Sep 2026 · github.com
Additional components
The project says it is based on ControlNet and BasicSR and uses Stable Diffusion v2.1 for training stage two.Read 12 Oct 2026 · github.com
Apple Silicon
The project announced Apple Silicon support and refers users to instructions for CPU, CUDA, and MPS devices.Read 12 Oct 2026 · github.com
Batch processing
YesRead 22 Sep 2026 · github.com
Maximum scale
8xRead 22 Sep 2026 · github.com
Model downloads
Pretrained weights are linked from Hugging Face, Baidu Netdisk, and OpenXLab, and the README says inference weights download automatically.Read 12 Oct 2026 · github.com
Model integrations
Inference uses BSRNet for super-resolution, SwinIR-Face for face restoration, and SCUNet-PSNR for denoising as stage-one models.Read 12 Oct 2026 · github.com
Online demo
The README announces an online demo on OpenXLab for the general and face models.Read 12 Oct 2026 · github.com
Quality trade-off
The README says using the condition as the sampling start point can improve stability but may decrease sample quality.Read 12 Oct 2026 · github.com
Restoration tasks
The repository provides inference examples for blind image super-resolution, aligned and unaligned face restoration, and blind image denoising.Read 12 Oct 2026 · github.com
Tiled sampling
Tiled sampling supports super-resolution at large scale factors on low-VRAM graphics cards.Read 12 Oct 2026 · github.com
Training limitation
The README states that the stage-two training script does not support a validation set.Read 12 Oct 2026 · github.com
Upscaling method
aiRead 22 Sep 2026 · github.com
Versions
The README lists DiffBIR v2 and v2.1, and says v2.1 includes a model trained on Unsplash captions generated with LLaVA, more samplers, and improved tiled-sampling support.Read 12 Oct 2026 · github.com

Plans and prices

DiffBIR Free Open-source research code and pretrained weights; GPU memory can limit inference Read 12 Oct 2026 · github.com

Every DiffBIR plan and price →

Where Specifiction places it

Alternatives to DiffBIRAll alternatives →