chaiNNer

chaiNNer is a free, open-source app for chaining image and video processing steps into custom workflows. You build each workflow from blocks called nodes, where every node does one job such as upscale, resize, or save. It started as an AI upscaling tool and grew into a workstation for many tasks. It runs on your own computer, so your files stay private and there are no fees. It supports popular AI models for upscaling, frame interpolation, and cleanup. Check the GitHub repo for the latest downloads and setup guides.

Key Features

1. Node-Based Workflows

Connect small processing blocks into a chain. You see every step of your workflow, and you can save chains to reuse them later.

2. AI Upscaling Models

Load popular upscaling models like ESRGAN and SwinIR. The app runs them on your own files with settings you control.

3. Video Processing

chaiNNer can process video frame by frame. Split a clip into frames, enhance each one, and join them back together.

4. Frame Interpolation

Add RIFE or FILM models to create smooth slow motion. They generate new frames between your real ones.

5. GPU Acceleration

Work runs on your graphics card when one is available. This makes large upscaling jobs finish much faster.

6. Local and Private

Everything runs on your own machine. Your images and videos never leave your computer, and there is no account needed.

7. Free and Open Source

The code is free under the GPL-3.0 license. Anyone can download it, use it, and help improve it.

Frequently Asked Questions (FAQ)

What can I build with chaiNNer?

Any image or video workflow you can chain together — upscaling, color changes, format conversion, frame interpolation, and batch jobs.

Does chaiNNer work on my computer?

It runs on Windows, macOS, and Linux. A graphics card speeds things up, but the CPU can run models too, just slower.

Is chaiNNer hard for beginners?

The node layout looks technical at first, but you can start from example workflows. Each node has a short built-in description.

Can chaiNNer upscale videos?

Yes. It splits a video into frames, runs your chosen model on each frame, and stitches them back into a clip.

How is chaiNNer different from a one-click upscaler?

One-click tools hide their steps. chaiNNer shows every step as a node, so you can change, repeat, and combine them freely.

Conclusion

chaiNNer suits people who want full control over AI upscaling and image work. The node system takes a little learning, but it replaces many single-purpose tools. It is free and open source, runs locally on your files, and supports the models researchers actually publish. Start with the example workflows to learn quickly. Batch nodes automate whole folders of images.