Skip to content
RepoQuarry

CodeFormer sczhou/CodeFormer

Face restoration model from NTU S-Lab with a fidelity slider, video input and face colorization; non-commercial licence.

CodeFormer at a glance

Checked 16 Oct 2026

CodeFormer restores blurry or damaged faces, with a weight (-w) that trades fidelity against quality, plus optional Real-ESRGAN background and face upscaling, video input, and colorization and inpainting for cropped faces. It is a Python research project (conda, PyTorch, CUDA). Code and weights are under the NTU S-Lab License 1.0, which allows non-commercial use only, so it cannot replace a paid tool for client work. It does not upscale general photos on its own. Last commit Nov 2025; weights released Aug 2022.

Cite this pageRepoQuarry. “CodeFormer.” Checked 16 Oct 2026. https://repoquarry.dev/projects/codeformer

How much of Topaz Gigapixel does it replace?

2 of 6 jobs. Strong face recovery for personal use only; whole-image upscaling relies on Real-ESRGAN. Command line, non-commercial.

Partial: Photo upscaling up to 16x Partial replacement
Partial: Face recovery on blurry or low-res faces Partial replacement
No: Recover and redefine low-quality or compressed images Not covered
No: Upscaling art Not covered
No: Batch upscaling of many images Not covered
No: Photoshop and Lightroom plugin Not covered

For recover and redefine low-quality or compressed images, upscaling art, batch upscaling of many images, 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.

Report a correction
QuestionAnswerSourceChecked
Runs on my machine? Confirmed: Yes, Python scripts on your machine python inference_codeformer.py -w 0.7 --input_path [image folder]|[image path] github.com · CodeFormer#readme ↗ today
Code licence No: NTU S-Lab License 1.0 (non-commercial) Redistribution and use for non-commercial purpose in source and binary forms, with or without modification, are permitted github.com · LICENSE ↗ today
Model weights licence No: Weights fall under the same S-Lab non-commercial licence This project is licensed under NTU S-Lab License 1.0. Redistribution and use should follow this license. github.com · CodeFormer#license ↗ today
Commercial use No: No, non-commercial licence for code and weights Redistribution and use for non-commercial purpose in source and binary forms, with or without modification, are permitted github.com · LICENSE ↗ today
Paid services needed? Confirmed: None required github.com · CodeFormer#readme ↗ today
Hardware Confirmed: PyTorch 1.7.1+ with CUDA 10.1+; VRAM not stated CUDA >= 10.1 github.com · CodeFormer#dependencies-and-installation ↗ today
Operating systems Partial: Not stated; the README has a note for Windows/Mac users (ffmpeg for video) For Windows/Mac users, please install ffmpeg first github.com · CodeFormer#readme ↗ today
Setup effort Confirmed: Medium, conda environment, pip requirements, download weights with a script conda create -n codeformer python=3.8 -y github.com · CodeFormer#dependencies-and-installation ↗ today

Get the weekly top 5 and what’s new, every Monday

We’ll email you once a week. Unsubscribe any time. We never share your email.

Cookie preferences

Choose what we may store in your browser. You can change this any time from Cookie settings in the footer.

Essential

Security, and remembering this choice. rq-consent · 12 months

Always on