Freedom251 › Image Upscaling Software
chaiNNer
What chaiNNer is
chaiNNer is a desktop node editor for assembling image-processing pipelines on Linux, macOS, and Windows. It can upscale images with AI by as much as 8x, resize, remove backgrounds, adjust color, blend images, convert formats, process batches, and work with video frames. For model execution, it supports PyTorch, NCNN, ONNX, and TensorRT, with CPU fallback. Its maintained Spandrel package supports ESRGAN, SPAN, OmniSR, and dozens of other architectures; OpenModelDB is offered as a place to find community-trained models.
A chain can branch conditionally, loop through folders or video, run math, and build text or filenames from values. Saved chains can also be launched from the command line as part of a larger pipeline. On first launch, chaiNNer downloads an isolated Python build and handles dependencies through its built-in Dependency Manager, though users can choose system Python instead. The maker calls chaiNNer a GPLv3 hobby project, with nothing gated behind payment. Discord is recommended for sharing chains and models and getting help, while GitHub issues are for bug reports. The Alpha v0.25.1 download page warns that macOS may show an installation warning because code signing was disabled to address build issues. Development is at reduced capacity, and ONNX DirectML support is currently broken.
Who it suits
chaiNNer suits people who want to build reusable image-processing workflows on a desktop across Linux, macOS, or Windows. Its branching, looping, batch, video-frame, and command-line options also fit users incorporating image processing into larger pipelines.
What it does well
- AI upscaling supports up to 8x scale.
- Batch processing and video frames are supported.
- Chains can branch, loop, and perform math.
- Saved chains can run from the command line.
- Free under a GPLv3 hobby-project model.
What to know first
- The Alpha v0.25.1 macOS download may trigger an installation warning.
- ONNX DirectML support is currently broken.
- Development is at reduced capacity.
Limits
The free plan gates nothing behind payment. ONNX DirectML support is currently broken; macOS Alpha v0.25.1 users may see an installation warning because code signing was disabled.
Plans
| Plan | Billed and limits | Price |
|---|---|---|
| Free | Nothing is gated behind payment · GPLv3 hobby project | Free |
As chainner.app lists them, read 11 Oct 2026. The maker's own page has the final word.
Getting started
- Visit https://chainner.app/ and choose the download for Linux, macOS, or Windows.
- Launch the desktop app; it downloads an isolated Python build on first launch.
- Use the built-in Dependency Manager, or choose system Python.
- Build a chain using supported image operations and models.
Verdict
Choose chaiNNer if you want a free desktop node editor for flexible image-processing chains, with model-engine options and command-line use. Look elsewhere if you rely on ONNX DirectML or need an actively staffed development pace.
Questions
How much does chaiNNer cost?
The Free plan costs nothing, and nothing is gated behind payment.
Which platforms does it support?
chaiNNer is available for Linux, macOS, and Windows.
Can it upscale images with AI?
Yes. Its maximum scale is 8x.
Is chaiNNer open source?
The maker describes it as a hobby project under the GPLv3 license.
What model engines can it use?
It runs models through PyTorch, NCNN, ONNX, or TensorRT, with CPU fallback.
Where can users get help?
The maker directs users to Discord for help and chain or model sharing, and GitHub issues for bug reports.
Facts from chaiNNer's own pages
- Batch processing
- Yes · chainner.app, 20 Sept 2026
- Command line
- Saved chains can be run from the command line as a step in a larger pipeline. · chainner.app, 11 Oct 2026
- Desktop app
- Yes · chainner.app, 20 Sept 2026
- Free plan
- Yes · chainner.app, 20 Sept 2026
- Image processing
- It supports upscaling, resizing, background removal, color work, blending, format conversion, batch runs, and video frames. · chainner.app, 11 Oct 2026
- Known issue
- The v0.25.0 release notes say ONNX DirectML support is currently broken. · chainner.app, 11 Oct 2026
- License
- The maker describes chaiNNer as a hobby project under the GPLv3 license. · chainner.app, 11 Oct 2026
- Maximum scale
- 8x · chainner.app, 20 Sept 2026
- Model engines
- It runs models through PyTorch, NCNN, ONNX, or TensorRT, with CPU fallback. · chainner.app, 11 Oct 2026
- Model support
- Its maintained Spandrel package supports ESRGAN, SPAN, OmniSR, and dozens of other architectures, and the site points users to OpenModelDB for community-trained models. · chainner.app, 11 Oct 2026
- Platforms
- The download page lists Windows installer and portable builds, a macOS installer, and Linux Debian/Ubuntu, Fedora/RHEL, and portable builds. · chainner.app, 11 Oct 2026
- Purpose
- chaiNNer is a desktop node editor for building image processing pipelines. · chainner.app, 11 Oct 2026
- Python setup
- On first launch, chaiNNer downloads an isolated Python build and manages dependencies through a built-in Dependency Manager; users can choose system Python instead. · chainner.app, 11 Oct 2026
- Release caveat
- The download page says the Alpha v0.25.1 release resumed after more than a year without a stable release and that development is at reduced capacity. · chainner.app, 11 Oct 2026
- Security note
- The download page says macOS code signing was disabled for Alpha v0.25.1 because it caused build issues and warns users they may see an installation warning. · chainner.app, 11 Oct 2026
- Support
- The maker directs users to Discord for sharing chains, models, and help, and to GitHub issues for bug reports. · chainner.app, 11 Oct 2026
- Upscaling method
- ai · chainner.app, 20 Sept 2026
- Workflow features
- Chains support conditional branching, looping over folders or video, math, and building text and filenames from values. · chainner.app, 11 Oct 2026
Where chaiNNer stands on Freedom251
- Image Upscaling Software#15 of 99