chaiNNer

chaiNNer is a desktop image processing editor for people who want AI upscaling and repeatable editing workflows without writing scripts. It runs models locally on Windows, macOS, and Linux. Connected nodes let you combine model processing with ordinary image edits in the same workflow.

Upscaling is its main focus. It supports community models from OpenModelDB, with Spandrel providing support for architectures such as ESRGAN, SPAN, and OmniSR. Face restoration and color correction can address faces and color shifts in an upscale. PyTorch, NCNN, ONNX, and TensorRT provide model execution, with GPU support covering NVIDIA, AMD, Intel, and Apple Silicon depending on the backend. CPU processing is also available.

The editor also handles background removal, color adjustments, blending, resizing, and format conversion. For game artists, it can generate and combine normal maps, convert PBR textures, save DDS files with mipmaps, and split spritesheets for processing before assembling them again. Pixel art tools include non-AI scaling, palette extraction, and dithering.

Batch workflows can process image folders or video frames, compare models, and skip an upscale when an image is already large enough. Outputs can use filenames built from image or model names. The editor checks connections for compatible types and marks errors on the responsible node; saved workflows can also run from the command line.

chaiNNer is open source under GPL-3.0. It downloads its own Python environment on first launch.