Open-Source Image Editor Ships New AI Engine

GIMP, one of the most widely used open-source image editors in the world, received an update incorporating a new AI-based processing engine, developed in partnership with researchers from the University of Grenoble. The new AI tools are available as a native plugin in GIMP 3.2 and include adaptive noise reduction, degraded image restoration, and object removal with intelligent background fill.

The New Tools in Detail

The AI Denoise Engine uses a model trained on millions of images with different types of noise to distinguish with high precision between noise and actual detail in the image — a problem that conventional noise reduction models solve imperfectly, often smoothing details that should be preserved. In comparative tests with paid tools like Topaz DeNoise AI, the GIMP 3.2 engine obtained competitive results on high-ISO photos from entry-level and mid-range cameras.

The Smart Inpainting tool allows selecting a region of the image and asking the system to replace it with content generated from the surrounding context. The current implementation works best on backgrounds with uniform texture — sky, grass, wall — than on complex scenes with multiple objects, but represents a considerable advance over the manual Clone Stamp tool that was the previous alternative in GIMP.

Impact for the Open-Source Community

The integration of AI in GIMP is relevant not just for the tools themselves, but for the signal it sends: features previously accessible only in expensive proprietary software are reaching the free ecosystem. The plugin is available under the GPL license and the AI model code is open-source, allowing the community to contribute improvements and adaptations. The update is available in GIMP's official repositories for Linux, macOS, and Windows.